How Do I Provide a Workstation To Last 15 Years?

An anonymous reader writes "My father is a veterinarian with a small private practice. He runs all his patient/client/financial administration on two simple workstations, linked with a network cable. The administration application is a simple DOS application backed by a database. Now the current systems, a Pentium 66mhz and a 486, both with 8MB of RAM and 500MB of hard drive space, are getting a bit long in the tooth. The 500MB harddrives are filling up, the installed software (Windows 95) is getting a bit flakey at times. My father has asked me to think about replacing the current setup. I do know a lot about computers, but my father would really like the new setup to last 10-15 years, just like the current one has. I just dont know where to begin thinking about that kind of systems lifetime. Do I buy, or build myself? How many spare parts should I keep in reserve? What will fail first, and how many years down the line will that happen?"

Comments

Moving parts are the main problem

[jonbryce]

Hard drives and fans will be the first to fail as they have moving parts.

You can get systems that don't need fans, but replacing the hard drives with flash memory probably isn't going to help reliability.

Re:Moving parts are the main problem

[Anonymous Coward]

Most of the failures of my home machines in the last six years have been fans and/or the power supplies housing them. (sometimes hard to tell what died first) With six desktop class machines running in the house, I've only had one drive failure, but I've replaced four power supplies and several frozen case fans. These aren't gaming rigs, just basic surf/email/homework boxes.

That said, with the price of used off lease gear on ebay and elsewhere these days, you could pick up machines that would run rings around the existing systems for uder $300.

Re:Moving parts are the main problem

[Shivani1141]

Not so with the flash drives. I was looking into an equivalent for MTBF for flash drives, and not finding one i started looking into the maximum capacity of writes, and found an article extolling a sort of half-life figure for flash drives. looking into the drive i have installed in my media center (for Quiet) an OCZ model. i found that i'd have to be writing to the drive at maximum capacity 24/7 for 18 years before the available capacity of the drive would decrease by half. they're quite long-lived, if the maximum writes per sector figures are to be believed.

Re:Moving parts are the main problem

[mariushm]

Build a computer with a processor that has very low frequency, something like an AMD Sempron LE-1300.

It runs by default at 2300Mhz but you should be able to lower it to something like 1Ghz or maybe even lower, which will increase the compatibility with DOS (if needed and if there are any incompatibility) and it also means that the computer will run even without the fan running over the processor.

You can solve the power supply fan problems by buying a passively cooler power supply.

You could also get a SSD drive or maybe a cheap Flash to IDE/SATA adapter and use 1 or 2 GB compact Flash card for DOS.

Though you can simply create a virtual machine or even DosBox (if you don't need some complex printing functions)..

Re:Moving parts are the main problem

[mysidia]

Keep in mind that it may be prudent to pick less-reliable hardware that should still last 4 or 5 years (most likely), over slightly more-reliable hardware, WHEN the price difference makes it more cost-effective to ANTICIPATE replacement.

Even the most reliable components may be expected to fail in 5 years.

I think he's been very fortunate that his setup has lasted 15 years. On average, a computer has had a lifetime of 5 years, before some hardware failures occur. To be honest.. in many cases, newer hard drives has been less reliable or has not lasted as long.

The higher data density results in more failures not less. The more bits (at essentially the same rate of defects), means it's much more probable for there to be at least one sector defective on a larger drive.

Power supplies can fail within 1 year or 10. It's random, so there can be no guarantee that the setup will last 15 years without any hardware replacement. (Even using the hardware he has right now, something could have failed in 1 year. A drive could go completely bad tomorrow.)

So get a very decent power supply, preferably one that is efficient at the anticipated load (which you should calculate for the chosen hardware), but can handle a lot more.

Using SSDs would improve reliability if used in a RAID 1 array, and a choice made with decent cache and wear-levelling, provided your app is reasonable they should last 50 years (typical use level), more likely the RAM dies first.

But unfortunately, the suitable SSDs of any reasonable size are also highly expensive. the cheaper ones don't have the few gigabytes or so of battery-backed RAM cache that would be necessary for high speed. --- Which come to think of it, may also be a reliability risk, since most types of rechargeable batteries don't last 15 years.

And I expect you don't need high speed for a small veterinary database, so the most inexpensive SLC or MLC may be just what the doctor ordered..

Another possible application for flash is simply to boot off of it, and then use an ordinary mechanical hard drive for storing your data. This way, mechanical wear is not introduced when you boot your OS, and writes are rarely required.

However, Windows XP (or Vista) is not suitable for this, as it likes to write to its own boot media. A Linux-based kiosk with a mysql-backed database app of some sort could work great there.

Make sure you get a lot more space than you need, i.e. try to fit everything you need within 5 or 6 GBs. And use a 50gb drive, so you can have an "active" partition and "backup" partition

Minimize mechanical wear on your drives by getting enough memory to run the workstations without a swapfile or pagefile. i.e. get 1GB or 2GB (a workstation that can use ECC memory is better, as you reduce the small possibility of silent data corruption), and make sure you disable all paging/swapping features within your OS.

Use the most reliable drives available for a reasonable cost; these are probably NOT 1TB 7200RPM drives; these are more likely 30gb 5000RPM drives that come with a 3 year or 5 year warranty.

Have each workstation backup the other workstation, i.e. so there are always two copies of the database. This is in addition to daily backups to external media to be stored offsite.

Unless you are using a UNIX/Linux OS with a journalled filesystem (or something like ZFS), it's pretty much a fact, that you are likely going to need an OS recovery at least once.

Each workstation should have two drives and a 'working partition' and backup partition. That you manually refresh every few months. Even better if they are separate physical disks (but again, more expensive)

Reliability will be maximized if you use a UNIX or Linux based application. And you minimize unnecessary reads and writes to your mechanical media, and minimize unnecessary load (and therefore heat) emitted by your hardware.

In any case, the usernames logging into the worksta

Re:Moving parts are the main problem

[ThePhilips]

Any chances that you still have the link(s)?

Because my reading of Anand's research tells me that in active, non-stop use SSD would fail in about the same time as normal laptop 1.8"/2.5" harddrives - 1-1.5 years. Limit on number of rewrite cycles is high (~100k), yet is quite easy to reach.

Re:Moving parts are the main problem

[Vadim Makarov]

Get well-designed fans? Might not worth the trouble for computers, but we get them for self-built scientific equipment with potentially long lab life.

Re:Moving parts are the main problem

[timeOday]

How about this system? Main thing for me would be to avoid fans at all cost. I hate fans. They crap out all the time, drag dirt into your system, make noise, and consume power.

Also, I would go with SSD drives. Intel makes theirs to last 5 years under heavy use, so I wouldn't be surprised to get 15 of light use.

Re:Moving parts are the main problem

[mmarlett]

Assuming that you have to rewrite his software, make it all web based (even if it runs off of one machine as a server without the Internet) and forgetaboutit. Keep it as basic and generic as possible and then the hardware will never matter.

Re:Moving parts are the main problem

[cbiltcliffe]

Well, since the old machines are 15 years old, and still running fine, my suggestion would be this:

Don't buy a new one. Don't build a new one. If you must have a backup computer, find another old machine that's in somebody's basement, garage, or otherwise not being used.
Build a low-power machine (Celeron, Sempron, whatever) with quality parts (3 year warranty, at least), with RAID, install Linux on it, and use it as storage for the database.
Pull the cover from the old machines, take an air compressor to them to clean them out, then replace all the fans. Again, use high quality parts.
Format and reinstall Windows on both, so the flakiness goes away. Install all updates, and the customer/patient management database, and configure it all to access the data on the server.
Then, pull the drive, and use something like Clonezilla on a laptop with a USB-IDE adapter to take an image of the drive and save it on the server.
Now you've got a couple of clean machines, with fresh software, redundancy for the data, and nobody has to deal with a change as drastic as Win95 to Vista.
If a drive fails, you've got an image of the software preconfigured.

After you've done this, keep an eye out for old drives in the 1-5GB range. Try to get at least 3 or 4 that work well, so you've got spares for when one fails.

As long as you don't get hit with a power surge or something, this is the most likely failure of anything this old, as it's just too low powered to generate enough heat to cause too many problems.
And if you need them, I've got a couple of AT type power supplies kicking around that work fine.

Also, make sure a proper backup is done of the data on the server. If he's got Internet access, encrypt it (GnuPG with a strong password or key) and send it to a gmail account, or something like that. Otherwise, a removable or USB drive that he can take offsite.

Re:Moving parts are the main problem

[nietsch]

But unfortunately, the suitable SSDs of any reasonable size are also highly expensive. the cheaper ones don't have the few gigabytes or so of battery-backed RAM cache that would be necessary for high speed. --- Which come to think of it, may also be a reliability risk, since most types of rechargeable batteries don't last 15 years.

ehm, Nand flash ram by itself is pretty fast, and Linux can natively handle it. Unfortunately, all available flash is slowed down behind some disk emulation chipset, which would make battery backed ram necessary. Just wait a few more months till the real good ssd come available... (as always in IT, things will be better later).

Re:Moving parts are the main problem

[Anonymous Coward]

No, the first thing to fail will his the pie-in-the-sky plan and it will happen as soon as the "new car smell" feeling of the system he builds wears off and he realizes that his expectations are unreasonable and that his question was inane.

Re:Moving parts are the main problem

[legoburner]

I've seen a SSD fail because the same sector - the superblock - was constantly overwritten to the point where it failed. Think how many times the superblock (if the relevant FS is using one) gets touched, then figure that in to reliability calculations too!

Re:Moving parts are the main problem

[jopsen]

Drive decreased by half. Is not a substainable solution... Drive decreased with only a few percentage for after that I'd replace it...
Also, won't you lose data?

Re:Moving parts are the main problem

[maxume]

Up to date controllers write heavily used logical sectors to new physical locations, so repeatedly writing a single sector isn't all that different from repeatedly writing other stuff.

Re:Moving parts are the main problem

[Ron Bennett]

Occasional thorough cleanings of dust off the fans and power supply will greatly extend the life of the entire computer.

Flaky / noisy fans, and especially, buzzing / irregular sounding power supplies are telltale signs of serious dust problems.

Sucking / blowing out dust is sufficient for most of the parts, but sometimes not enough for fans (including those on graphics cards, etc) and power supply where dust can easily cake up necessitating physical disassembly for cleaning.

In short, the OP, if they haven't already, should make regular dust cleanings part of the maintenance routine - that alone will add much life to the hardware.

Ron

Re:Moving parts are the main problem

[Ost99]

Failed sectors are supposed to still be readable. The controller on the disk just sets them as read only when they fail.

Re:Moving parts are the main problem

[Anonymous Coward]

Flash memory sectors tend to fail on write, and in a detectable way. So the internals just write to a different sector and mark the old one as dead. (The old one is still readable, too). You don't lose data unless you had already run out of space (including the spare sectors that were never exposed to you in the first place. Normal magnetic drives have extra sectors too for the same reason...).

Since the old computers are running 500 MB drives and today's flash drives tend to be 8GB or larger, he should be fine. Say 2 GB for XP and apps, and the 500 MB of old data, and he'll still be under 4 GB used before the drive hits its half-life in 18 years. I doubt it'll go that far without some kind of hardware fixes anyway. (the 15 years of his existing system already makes it an extreme outlier, IMO).

Re:Moving parts are the main problem

[Shivani1141]

Ah, found one of my original articles, oddly the one corroborating it 404s now. please, read it and make your own conclusions. http://www.storagesearch.com/ssdmyths-endurance.html

Re:Moving parts are the main problem

[Yoozer]

Pull the cover from the old machines

This may already be enough. It's not like he's going to actively look at the insides or right-click on My Computer to see that the drive has suddenly grown from 20 to 40 gb or something, so just replace the parts and use the cheap stuff when he's not looking or on a holiday. Least you could do for him, won't break your wallet.

Re:Moving parts are the main problem

[Anonymous Coward]

Have you ever thought about that claim? The wear-leveling claim that is? Because in my world that kind of behavior is going to need some sort of index so that the SSD controller knows where it is currently storing block 0 on the flash device. And that index must be non-volatile. So what kind of storage might be used for that? Hmm. Perhaps flash. And how often will that index have to be written to? Any time any other piece of the flash is written. (Even if the controller isn't moving the storage around on the device it still needs to know how often the current cell has been written.) So how many write cycles does that index data have?

Re:Moving parts are the main problem

[nemesisrocks]

Unfortunately, the mere act of cleaning out the dust in the machine might spell its demise.

I can't even count the number of times I've thought "hey, there's a lot of dust in that machine, let's clean it", and the machine refused to power back on afterwards...

Re:Moving parts are the main problem

[usasma]

The critical portion of this is the veterinary software. That will control what you can build. I upgraded my wife's veterinary office from a DOS veterinary software to a Windows veterinary software several years ago. The software isn't readily compatible with Vista, so we've remained with XP for most of the workstations (6 workstations, 1 file server). About a year ago I upgraded a Windows 2000 server network to a Windows 2003R2 server network - and it was relatively painless because of the compatibility of the software. If you do the work, you become the IT department for the practice. It's not much of a hassle, but it does require that you drop what you're doing to fix their computers (I got a call recently to rush over to the practice - only to replace a keyboard and run back to my primary job). OTOH it's more expensive to have an outsider do the work - but it's easier for you to supervise the work. The backup strategy is the key here (IMO). If you've got good backups you can start over at the drop of a hat. I'm starting to experiment with RAID 1 for the practice, and I backup the file server 3 ways (to another drive, to another computer, and to another location).

Re:Moving parts are the main problem

[unitron]

Pull the cover from the old machines, take an air compressor to them to clean them out...

Make that compressor a vacuum cleaner. You want to remove dust and crud and stuff, not drive it in deeper.

Re:Moving parts are the main problem

[maxume]

So allocate some conservative number of sectors for the index (conservative in the sense that it should outlast the rest of the drive) and look through those sectors in the opposite order that they are used until you find the most recently used sector, which is the up to date index.

If I can come up with that in 10 minutes (the only flaws would be that it might take some milliseconds at start-up, and it is subject to poor estimation), Intel can come up with something that works just fine.

Re:Moving parts are the main problem

[ThePhilips]

Thanks for the link. Quote:

I found some data from Mtron (one of the few SSD oems who do quote endurance in a way that non specialists can understand). In the data sheet for their 32G product - which incidentally has 5 million cycles write endurance - they quote the write endurance for the disk as "greater than 85 years assuming 100G / day erase/write cycles" - which involves overwriting the disk 3 times a day.

The info though seems to be usual fine PR. What I see that for example Samsung quotes for 2.5" HDD MTFB of "less than 700,000 hours" (~ 80 years!!!) and for 2.4" SSD MTFB of "more than 1,000,000 hours" (~ 117 years).

Have you *ever* seen HDD surviving 80 years? Nope. (Ask any SAN admin for references.)

In other words, I'm not buying the "greater than 85 years" mark from Mtron. Equally I do not buy the numbers from Samsung.

Though in the end, it sound a bit reassuring that in that respect the SSDs at least not worse than HDDs.

Re:Moving parts are the main problem

[mysidia]

Yes, but they're going to be so expensive you'd be better off buying a pair of the cheap old-generation 20gb IDE (or SATA) hard drives.

Put the difference between the cost of your drive and a SSD in the bank (>$200). When one of your mechanical drives eventually fails, take that workstation out of service, and just use the other one to get by when that happens.

Use the money in the bank to acquire a second mechanical drive, or more likely that GOOD SSD, which in 5 years, will be dirt cheap and pretty reliable.

Re:Moving parts are the main problem

[Anonymous Coward]

Though in the end, it sound a bit reassuring that in that respect the SSDs at least not worse than HDDs.

Right, which was my original point as well. It's important to note that i'm not saying that SSDs last significantly longer than a Mechanical drive, they just arent worse, which seems to be the popular misconception.

In reference to the original parent poster my point stands, an SSD would be more reliable in this application than a magnetic drive, if for not other reason than it's not going to suffer mechanical failure in those fifteen years.

Re:Moving parts are the main problem

[Anonymous Coward]

The figures might be calculated differently but the ridiculously large MTBF figures they calculate for regular HDDs are based on replacing the unit every few years (based on the warranty lifetime). Might want to double check for this.

Re:Moving parts are the main problem

[davolfman]

That might just be the difference between SLC and MLC flash.

Re:Moving parts are the main problem

[davolfman]

Wouldn't that make the best approach to virtualize the entire OS and application so it can be migrated or restored to new hardware at a moments notice? That is one of the big selling points of virtualization isn't it?

Re:Moving parts are the main problem

[N3Bruce]

I too have to keep old DOS based systems running at work on the host PCs on our mail extractors. Case in point: A machine we originally got in 1996 came with a cheap Acer 486/33 processor with 8 megs of RAM. I ran that system for 9 years, defragging and running Scandisk on the hard drive and blowing the dust out a couple of times a year. In that time, the only repair I had to make to the system was to replace the CMOS battery after about 6 or 7 years. A power supply failed, I got a used one and forgot to set the voltage selector to 220 volts. Well, that was the end of the Acer.

I got a replacement PC from HQ that was a Celeron 533 or something similar, and it lasted a year or two. Fan crapped out on the CPU. Got a replacement that was a 2.4 Ghz Celeron that lasted only 3 months- Again, fan crapped out on the CPU. If I only hadn't screwed up the Acer, it would probably still be running today.

BTW, I have a Pentium 100 running a sorter that is still going strong going on 10 years. My main problem with the older PCs is backing them up. Floppies are slow, and have become so unreliable that they are hardly worth the bother. Running Ghost is an option, if the Ghost disks on floppy can be read that is.

Re:Moving parts are the main problem

[Glonoinha]

Have you *ever* seen HDD surviving 80 years? Nope. (Ask any SAN admin for references.)

That's not how MTBF works. It's an aggregate across the entire enterprise. Let's say you populate your infrastructure with 1,000 2.4" SSD's with a MTBF of 1,000,000 hours. In theory, you can assume that you're going to have one drive fail every 1,000 hours (or roughly one failure every 6 weeks, or roughly 9 failed drives each year.)

Re:Moving parts are the main problem

[mysidia]

Sure, by all means. If you have the skills to do it, virtualization is great. The problem is the barriers (human difficulty) of getting it successfully deployed and setup to work flawlessly.

Make a bootable IDE flash drive running OpenSolaris, script it to start up a desktop hypervisor and load a virtual machine image with a Virtual Disk File stored on the hard drive, and use a script to boot the virtual machine and make it full screen whenever the system starts up.

Have a script running in the backround that periodically takes a ZFS snapshot of the pool, and periodically synchronizes the state of the second disk on the other workstation (using zfs send over SSH).

You can easily and inexpensively make a few copies of this IDE flash drive.

Re:Moving parts are the main problem

[Darkness404]

Put the difference between the cost of your drive and a SSD in the bank (>$200). When one of your mechanical drives eventually fails, take that workstation out of service, and just use the other one to get by when that happens.

...But you can easily get a $100 SSD (8 gigs) that based on the post, would be enough to run the system (either XP, or Linux with DOSbox or other emulation software), and save the documents for about 15 years. This person isn't going to be doing anything more disk intensive as saving about 5 documents a day, assuming that this is computer will always be left on also decreases SSD wear, add in the fact that SSDs fail gracefully (as in, you can still access all the data, just not write it) compared to the catastrophic crashes of HDs alone makes SSDs a good option (considering most people don't make backups). Its easy to make a good system for ~$400-500 with high quality fans, a SSD, and a low-voltage CPU that should last 15 years.

Re:Moving parts are the main problem

[AchilleTalon]

He may be fortunate his setup last so long. However, I personnally have a PC 486 I bought in 1993 and it is still running 24/7 without a glitch. The only problem I had so far was with the CPU fan I changed about 6-7 years ago. I also have a Pentium 66 running also without a glitch since 1995. This last one has its CPU fan replace, the bracket has broken and I drilled a hole in the mother board to tied the CPU fan and heat sink to the CPU using a iron wire I passed throught the hole and twisted (I did this more than 10 years ago). The other problem I had, the PSU fan died and then the PSU itself died and was replaced about 2 years ago. Beside this, I have no problem with these systems running Linux. The 486 is having 16 MB RAM and the P66 128 MB RAM.

I am seeking to replace my 486 by a Soekris SBC I already have and am in process to build the Linux code for. The P66 will also be eventually replaced by another Soekris SBC after my first one has been done successfully. So, I will retired my old PCs not because they are broken, but only because I can not longer afford the risk of a failure since it is becoming harder to find replacement parts and also because the 486 is a little bit short in RAM for what I want to do with it.

Re:Moving parts are the main problem

[Anpheus]

I'm going to go ahead and throw out the idea that maybe the people who develop these SSDs are smart enough to realize that.

Re:Moving parts are the main problem

[Anonymous Coward]

Modern flash drives do block rotation at the hardware level.

Re:Moving parts are the main problem

[huded]

If he is using Win95, M$ stopped supporting that years ago; I'm pretty sure there are no Windows updates for that OS on their site anymore.

Re:Moving parts are the main problem

[DrgnDancer]

Oh come on, Virtualization isn't that hard. It's not like he's trying to build a huge virtualized computational cluster or something. Install RHEL 5 with the Zen kernel extensions. You can use the GUI to do most of the rest. Get a nice workstation with a couple of dual core CPUs, 4 gigs or so of RAM and a couple of 500G drives. Build however many VMs as you need (2 to replicate the old system if that's easiest), and snapshot them all regularly. Backup to to the second drive. If you really wanna be paranoid, backup to an external drive and keep it in a fireproof safe. Whole rest of the building could burn down and with a new workstation you could have the computer system back up and running inside 4 hours. With a not terribly expensive 2 x dual core system he's could totally flub the VM resources and not even notice compare to what they had before. Domain zero could use half the resources on the computer and the VMs would still seem down right snappy.

Re:Moving parts are the main problem

[lorddarthpaul]

I think its unrealistic to expect a consumer level computer system to assuredly last more than 5 years (isn't that the standard depreciation schedule anyway?), especially if you plan to add additional software to it over time (newer software tends to be ever more demanding?). My strategy is to buy cheap (well, if it's not Apple ;-), but reliable (i.e. read reviews) hardware and replace parts, if necessary, in the interim. Buying reliable does not have to mean buying expensive.

As far as drives go, I have a 1TB WD10EACS being beaten to death daily in a TiVo HD (since May 2008) and it's pretty reliable so far, as are all but one of my 200GB ATA drives (works once it starts, but has a flaky spin up -- I suppose I should stop using it?). Many of my older SCSI drives (4GB, 8GB, circa 1995) seem to work as well as they ever did -- I've sold some to agencies looking for replacements in nuclear power plants and similar applications (don't you feel safe now?). However, an 8GB USB flash key I was using became completely inert one day -- and it didn't seem to be a static shock issue. I had an identical 1GB one that worked fine for years, and I went back to that until I got another 8GB one (different brand this time). Didn't lose anything, as it was only being used as a backup device. For unchanging applications that never need updated software, I suppose you could just keep an old system going forever, maybe backing it up just enough (clone drive?) so you could buy a replacement old system on ebay and swap it in if necessary. Works, but it's generally better to "get modern" when possible --- though it never really stays that way for long!

Re:Moving parts are the main problem

[Anonymous Coward]

and the tape drive, which you should do at least weekly backups with as even mirrored (or raided) hard drive can be corrupted.
Since he doesn't need much. I would look for something cheap like a HP Proliant DL360 G3 or G4 or a DELL PE1450 and try to image it over to it.
Or two cheap Dell workstations with network backups. The HP Proliant systems are lasting a long time so far.

Re:Moving parts are the main problem

[Anonymous Coward]

Under-clocking is an old trick that works. Glad to see someone else remember and mention it here on allthingsshiny.com

Re:Moving parts are the main problem

[Anonymous Coward]

You're either full of shit or talking about a very very old flash drive without any wear leveling whatsoever. Current drives do not write to the same block of flash memory, even if you keep writing to the same logical block.

In regard to the original question: Don't worry too much about longevity. New computers are much cheaper than they were 15 years ago. It is unreasonable to pay significantly more to get a machine which outlasts a cheap PC. Failure still isn't among the top reasons for replacing computers. Usually they're being replaced for the reasons the submitter stated: It needs upgrades which are no longer on the market, the user desires modern software which needs a faster machine or a different OS with incompatible drivers, etc. In other words, the computer still does what it did, but that's no longer enough. Buy a new computer now and your chances of replacing it due to breakage sometime in the future are still very slim, despite the price crash.

Re:Moving parts are the main problem

[mdwebster]

You could also put a small EEPROM on the controller to store that information. A 128-bit EEPROM would probably be less than a dime in large quantity. A 1k-bit would cost just a penny or two more.

Re:Moving parts are the main problem

[mdwebster]

It's Mean Time Between Failure. Your example gives it as a hard number that none of the drives will survive past. There will be a statistical spread, but not like you describe it.

The actual spread depends on the shape of the failure distribution which isn't normally given. In a lot of cases, it's a Weibull curve which front-loads the failures and has a long tail. Even then though, you don't expect more than 60 to 70% of drives to have failed by the MTBF point.

Re:Moving parts are the main problem

[Lehk228]

you could also use a small chunk of battery backed up SRAM that mirrored itself to a block of flash every hour or so and on shutdown.

Re:Moving parts are the main problem

[maxume]

The trick is knowing which block of flash is current (while not wearing out that block of flash).

Re:Moving parts are the main problem

[Anonymous Coward]

Flash ages in a peculiar way: A flash cell is naturally in either 0 or 1 state (let's assume 1). You can flip a bit to 0 and it doesn't age the chip at all. You can not flip it back to 1 though. That "erase" step can only be applied to a whole block of flash cells and this step does age the chip. That's why flash memory durability is specified as a number of "erase cycles".

You can work with that to radically reduce the number of erase cycles needed to keep a logical-physical mapping. For example, you can add mappings to the end of a list (no erase cycle needed since you're just flipping some bits from 1 to 0) and mark outdated mappings by setting all bits to zero (again, no erase cycle needed). The controller just needs to look for the last mapping which isn't all 1s to find the end of the list and ignore mappings which are all 0s (they've been "deleted"). It can keep a sorted copy of the mappings in internal RAM to increase the lookup performance. Only when the list is about to overflow its allocated memory do you need to erase it and write a condensed version back. (You would use two lists and mark the latest with an incrementing serial number when it has been written completely, to avoid losing the mapping when the power goes out while you compact the list.)

Actual wear leveling algorithms are proprietary, but you see that wear leveling is certainly possible and does not necessarily need other, more durable memory to work.

Re:Moving parts are the main problem

[Anonymous Coward]

The leveling score for each block is written within the block and not in a separate "index" area.

Re:Moving parts are the main problem

[hedwards]

Or something from Soekris engineering, it's really low processing power, but if a text prompt is fine, there's no need to go with anything much more powerful. With the benefit of not having to worry a whole lot about.

I know some of the newer versions come in a 500mhz speed. Plus it runs both Linux and FreeBSD so you don't have to worry much about which version of Debian to use. ;)
If the application is that low powered, there really isn't any reason to get a 1ghz based computer.

Re:Moving parts are the main problem

[djlowe]

I too have to keep old DOS based systems running at work on the host PCs on our mail extractors

Were I in your situation I'd pick a VM solution and virtualize the application on modern hardware. You can do it and test it while the current hardware is still in service, and, once proven, you won't have to worry about replacement hardware: Any hardware that can run the VM software will suffice without having to twiddle around with the DOS VM once you have it running properly.

Of course, I'm making quite a few assumptions about your environment: You don't mention how the current systems interface to the mail extractors, for example - that could be a problem. Also, there's the potential issue of network connectivity. The fact that you mention backing up to floppies implies that you don't have any other means to do so... but, with a modern system running the DOS application in a VM, you'd have the option of backing up the DOS VM to external USB hard drives, or perhaps using PCI WiFi cards to get network connectivity and do it that way.

Anyway, all of the major VM applications suitable for such a test are free, it'd certainly be worth a try, if only to see if it were feasible.

Regards,

dj

Re:Moving parts are the main problem

[fractoid]

I believe EEPROMs use the same basic storage mechanism as Flash memory; the difference being that Flash can only be erased blockwise rather than bytewise, and consequently allows higher memory density due to fewer erase circuits. I know EEPROMs that I've used in the past have much lower erase limits than the numbers quoted for Flash these days - usually between 100 and 10,000 erase cycles.

I'd suggest having a few KB of flash ('index sector flash', ISF) set aside for storing the address of the current location of the index. When erased, the whole ISF would be zeroed, and every time the index is written to a new sector, the sector is appended to the ISF. The current location of the index is the last non-zero sector ID in the ISF. This way the ISF only needs erasing every few thousand writes, and the system scales easily to match the ISF lifetime to the wear-levelled life of the rest of the drive.

Re:Moving parts are the main problem

[Protonk]

Eh....kind of. MTBF comes from the assumption of an exponential distribution for failure in devices. Depending on whether or not failures of multiple devices are independent of each other, the probability of failure for a bunch of them may be normally distributed. Or it may not.

Re:Moving parts are the main problem

[fractoid]

Let's say you populate your infrastructure with 1,000 2.4" SSD's with a MTBF of 1,000,000 hours. In theory, you can assume that you're going to have one drive fail every 1,000 hours.

Erm, no. For starters, drive failures aren't uniformly distributed between installation and some arbitrary end-of-life cutoff. All you can say in your example is that if you sum up the total life of each of those 1000 SSDs and then divide by 1000, you'll get ~1000000.

What the GP was saying was that the quoted MTBF was patently ridiculous. Most hard drives will die long before they're 10 years old, even in home usage let alone server drives that get pounded 24/7. 80 years is at least an order of magnitude too optimistic.

Re:Moving parts are the main problem

[fractoid]

True, but I think the vet in question wants the actual box to last 15 years, so the tips above are useful too.

I'd just add that you should try to stay away from anything with electrolytic capacitors on it. They're usually the first thing to go - these days some motherboards are advertised as "no electro caps".

Re:Moving parts are the main problem

[ponraul]

"The only problem I had so far was with the CPU fan."

How did you notice the fan? Most of the chipsets that I remember from that period didn't have APM/ACPI support that included CPU temperature thermal sensors. Did you notice the case getting hot or did you notice your applications behaving erratically?

Re:Moving parts are the main problem

[myspace-cn]

FANS!!
I went to a local pet store, (I know not the same as a vet) and I was looking at the back of their computer.

The fan was packed with FUR! (And I mean PACKED!!!)

It dawned on me, how does this machine continue to run? Then I walked around the store a bit. I noticed several unplugged workstations stacked like they were being stored.

EACH had FUR PACKED IN THE HOLES, FANS, etc..
I had my answer. They die.

Re:Moving parts are the main problem

[Geminii]

Alternative: POSIX compliance. Runs on free OSen; many modern versions of Windows; likely to be able to easily migrate to updated POSIX-compliant platform in 2024.

Re:Moving parts are the main problem

[Anonymous Coward]

The new SLC drives are much more durable, but much more expensive. Expect to pay near $1000 for a 32GB drive. Put 3 of them in RAID 5, and get 2 spares. Make sure to have an audible or otherwise noticeable alert when a drive dies. That is $5000 for a relatively small amount of storage, but I think it might last a decade.

Re:Moving parts are the main problem

[mysidia]

$100 for an 8gb SSD is a lot, when you can get a 80gb SATA drive for $30. You can buy 3 such mechanical drives for less than $100.

The long term failure characteristics for SSDs are not well understood, but there are failure modes that can result in loss of data. Write cycle isn't the only failure possibility.

It seems unlikely they're just replacing the workstations to do a hardware refresh; a switch of OS, and use cases that will use a lot more disk space are likely.

8gb is pretty limited for a workstation. Especially considering their 500mb drive is filling up with files, and they are running Windows '95, and not Vista + Office 2007

Windows XP could utilize 5gb after a few years, for just the OS, after updates alone, with no Office install. Vista will utilize almost that much out of the box, and they may have to pick Vista, since they want the install to last so long (And XP won't be supported in terms of security updates for very many more years)

Add to that the approximately 500mb they are already using + additional space required to save the same documents using new software.

And their storage requirements will be ballooning. If they used 500mb in 15 years, one might guess they would use close to an additional 500mb in another 15 years, without a change of OS. It is probable that 8gb will not be sufficient for a full 15 more years, or that it will be a limit that burns or impacts them.

Maybe they'll never update it, and they'll use only small software with very sparse storage requirements, and a very compact, small database, but storage shortage could be a risk, esp. if E-mail becomes a desired capability of the workstations, (then Windows updates for Windows workstations becomes an essential thing).

And 5 documents a day may not be that warranted assumption. They might be scanning things that result in images, and expect the system to store them, over 15 years, this can add up to several GBs.

They may have some soft needs that the current system doesn't meet, and the new one ought to.

It's worked fine for 15 years, they have to be thinking about replacing it for a reason.

For a reason better than anticipated failure And windows '95's perceived flakiness.

If windows '95's flakiness hasn't been a major issue for 15 years, then they could probably survive another 15, do a simple software refresh, etc.

Re:Moving parts are the main problem

[David Jao]

Any chances that you still have the link(s)?

Because my reading of Anand's research tells me that in active, non-stop use SSD would fail in about the same time as normal laptop 1.8"/2.5" harddrives - 1-1.5 years. Limit on number of rewrite cycles is high (~100k), yet is quite easy to reach.

The article you cite does not contain the 1-1.5 years figure anywhere. How did you get that number? For what it's worth, I've been using solid state drives in both my laptops for more than a year now, with no problems whatsoever.

Another very important point which often gets ignored is that a solid state drive failure is far more benign than a spinning platter drive failure. When a solid state drive fails, you lose the ability to write data, but you can still read data. On the other hand, failure of a spinning platter drive means that you can't read your data anymore, at least not without sending it to a very expensive data recovery firm.

Re:Moving parts are the main problem

[mysidia]

RHEL 5 has no good snapshotting capability; I wouldn't wish LVM snapshots on my worst enemy, due to their serious performance penalty, and metadata corruption issues, and "snapshot running out of space" issues (Snapshot vanishes before your very eyes). LVM operations can fail in horrible ways you might not imagine, and they DO... especially if the plug gets pulled on the computer or the OS crashes in the middle of a snapshot, or before a snapshot is properly closed (after the copy off operation).

It would take an incredible development effort to gracefully recover without intervention from an expert.

That's why (until btrfs comes out and becomes stable); Opensolaris is the only good choice I know if you want a snapshotting feature on a free operating system that is robust and safe to use.

Building images is easy, the issue is making it seamless for the clueless user, and making sure the virtualization environment auto recovers from just about any failure (like VM instance crashing).

Re:Moving parts are the main problem

[packeteer]

You are right that fans and power supplies usually go first. I personally use a little bit of an overpowered power supply so it runs cooler and more stable.

Also I think the more general concern I have is that possibly planning for a 15 year lifespan might be the wrong way of looking into this. It is always much better to have a flexible upgrade and repair plan than try and force something to last much longer than it is intended. Make no mistake that consumer hardware is not intended to last 15 years.

I would much rather look at something like software of a data base that can upgrade smoothly in the future.

Re:Moving parts are the main problem

[TheKidWho]

If anyone around has seen an 84 year old HDD they would probablly be too busy telling the damned kids to get off the damned lawn.

Re:Moving parts are the main problem

[tsm_sf]

Assuming that you have to rewrite his software, make it all web based (even if it runs off of one machine as a server without the Internet) and forgetaboutit. Keep it as basic and generic as possible and then the hardware will never matter.

...and then pick a host or service that will be around in a decade or two.

Re:Moving parts are the main problem

[Sj0]

I'd say an online UPS as a component to help prevent premature power supply failure. It rectifies the signal at all times and creates a new perfect sinewave at all times. That'll get rid of transients and make your power supply far more reliable after you get past infant mortality.

My full solution would be a fanless rig, with RAID 1 for full redundancy of disks so if a hard disk fails, it doesn't take your data with it, and weekly backups to DAT tape stored off-site. Then I'd use a pair of power supplies, using a diode to prevent power from one from getting into the other, and a zener diode or 78 series linear regulators to ensure a failing supply can't overpower any one line. Then, from my little power circuit, the two power supplies would feed the one motherboard, which would be underclocked at reduced voltage. It would have the highest possible amount of RAM in it, because that would reduce the writes to the hard drives.

That should be reasonably reliable.

Re:Moving parts are the main problem

[Anonymous Coward]

he's giving you an aggregate replacement rate, dumbass

look up the downing effect, it applies to you

Re:Moving parts are the main problem

[oracleguy01]

Even the most reliable components may be expected to fail in 5 years.

I think he's been very fortunate that his setup has lasted 15 years. On average, a computer has had a lifetime of 5 years, before some hardware failures occur. To be honest.. in many cases, newer hard drives has been less reliable or has not lasted as long.

I think a big part of the reason those computers have lasted so long is that they produce so much less heat than computers these days. Heat is a big killer over the long term to electronics. They don't require so many heatsinks in the systems and they were probably built with better parts. Less hardware density also means less things to go wrong.

Re:Moving parts are the main problem

[InfoJunkie777]

I have had some kind of computer since 1988. The results so far: Power supply - 3 Motherboard fried - 2 Hard drive failed - 1

Re:Moving parts are the main problem

[Anonymous Coward]

How dare you, everybody here knows that the Slashdot comment database is really a vast repository of solutions on any of the current problems facing the world today, including (but not limited to) the popper design of an SSD disk, global warming, warp drive design, how to fix the LHC, getting to and back from Mars, string theory, correct grammar, how to make profit out of anything, how things would have been done in Soviet Russia and the current economic situation.

Thank you very much...

Re:Moving parts are the main problem

[basicio]

Maybe, but part of the goal is also to not fry the motherboard with static.

Which I had a friend do when he vacuumed his computer out.

Re:Moving parts are the main problem

[ShakaUVM]

>>Keep in mind that it may be prudent to pick less-reliable hardware that should still last 4 or 5 years (most likely), over slightly more-reliable hardware

What is reliable hardware?

I'm sure every nerd on here has his favorite brand of motherboard and hard drive, but by and large, we don't have the slightest idea which DVD drive out nowadays will have a high or low failure rate 5 years from now.

We like to think we know more than we do.

Re:Moving parts are the main problem

[hvdh]

MTBF is defined quite unintuitive. It does not say anything about a single drive, but average failure of a *lot* of drives which are only used in their specified lifetime.

So, you take some thousand drives and let them work. If a drive fails, you replace it and count a failure for the MTBF statistics. If a drive works to the end of its specified lifetime (e.g. five years), you also replace it, but it does *not* count as a defect for MTBF.

Then, you calculate (number of drives) times (hours of service) and divide by the number of defects replaced, and this is the MTBF.

Re:Moving parts are the main problem

[Anonymous Coward]

"non-stop"

I highly doubt your laptops are used as such.

Re:Moving parts are the main problem

[palegray.net]

and then pick a host or service that will be around in a decade or two

I'd like to think my favorite host (Linode.com, for those interested in quality VPS hosting) will be around in a decade or two, but I can't even be sure the term "hosting" will mean the same thing by then.

Re:Moving parts are the main problem

[supernova_hq]

Have you ever tried vacuuming a computer?!?

From a physics perspective, compressed air has a LOT more energy than a vacuum. Remember, you can only have so much of a vacuum, but compression is infinite.

I have tried vacuuming a system (suction) and it does shit all. I then hooked that hose to the output of the SAME vacuum (pressure), and you wouldn't believe the amount of dust that came out of that thing.

Just remember, DO IT OUTSIDE! :p

Re:Moving parts are the main problem

[Anonymous Coward]

original submitter here

The systems pretty much still work fine, but anticipated hardware failure is definatly a factor. So far the only hardware that needed actual replacement where:
- several CRT screens, they croke about every 4 years
- several cpu fans on the pentium, it uses small 40 mm fans, eventually the axel/barings gives out
- an optical drive

These were all low-impact problem, and spares werent hard to find, but if a one of the hard drives dies, or a cpu gives up the ghost, then that system might be permanently out by now.

As for system requirements, we havent done a full use-case analysis yet, but i dont think we will need much more functionality then we have now.

As for the 5 documents a day, that doesnt fly, my dad may not edit lots of documents, but his management software uses a database system with which i am not familiar, for all i know it updates data every minute..

as for email/net needs, i plan on keeping the entire system of the internet, as it is right now. Currently the systems have no use-cases that require internet (there is a third, much more up-to-date system which has a net connection, but no business-vital functions), and from what i've seen in the past years, complete isolation is rather handy when you want a reliable system

Re:Moving parts are the main problem

[ThePhilips]

Yep, I've read that.

Apparently, storage manufacturers include the MTBF because it is TOTALLY unrelated to real life. And no doubt it looks good in ads.

Re:Moving parts are the main problem

[Techmeology]

Then he needs an operating system with an automatically shrinking file system.

Re:Moving parts are the main problem

[yahwotqa]

Assuming that you have to rewrite his software, make it all web based (even if it runs off of one machine as a server without the Internet) and forgetaboutit. Keep it as basic and generic as possible and then the hardware will never matter. ...and then pick a host or service that will be around in a decade or two.

...and then read GP post again, more carefully, to find out why that is not necessary

Re:Moving parts are the main problem

[neurosine]

They're not a problem if you have a NBD warranty which can be extended at the end of the warranty lifecycle. They're not even an issue.

Re:Moving parts are the main problem

[CarpetShark]

What? In which fantasy world does setting up a WEBSITE (of all things) and ignoring it mean it'll still be running in 15 years?

Websites on the net require constant security upgrades, have a higher load even from random hits, are slow, aren't guaranteed to work on future browsers, and in this situation, you're also doubling the number of systems to maintain by moving to a client/server architecture.

The only sane way to handle this is to say "sorry, you got lucky with the past 15 years. But I can sell you the machine, and a 15 year support contract/guarantee."

Re:Moving parts are the main problem

[doodleboy]

My full solution would be a fanless rig, with RAID 1 for full redundancy of disks so if a hard disk fails, it doesn't take your data with it, and weekly backups to DAT tape stored off-site. Then I'd use a pair of power supplies, using a diode to prevent power from one from getting into the other, and a zener diode or 78 series linear regulators to ensure a failing supply can't overpower any one line. Then, from my little power circuit, the two power supplies would feed the one motherboard, which would be underclocked at reduced voltage. It would have the highest possible amount of RAM in it, because that would reduce the writes to the hard drives.

On the software side, I would consider hosting the DOS app on linux using an emulator such as dosemu or dosbox. The OP's dad would have an environment very similar to what he's using now. I would probably use Debian stable for both boxes, which has very long release cycles and is very stable.

With linux comes the option to replace the DAT tapes with an off-site rsync over ssh. If the main box dies, you'd be able to just swap in the backup box in a couple of minutes. If the data set isn't very large the mirror will complete in a couple of seconds. It's very easy to do:

Create a RSA public/private key pair: ssh-keygen -t rsa, press enter at the password prompts.

Copy the public key to the remote box: ssh-copy-id -i ~/.ssh/id_rsa.pub remotebox.

Have a nightly cron job to push the files: rsync -ave ssh --delete /localfiles/ remotebox:/localfiles.

For bonux points you could even throw in snapshots.

I'm backing up hundreds of partitions this way at work, each with snapshots going back a month. Tapes are slow, unreliable and expensive. I would not use them for any purpose.

Re:Moving parts are the main problem

[Anonymous Coward]

If you are using the same DOS software this is not an issue as writes will be very infrequant.

Re:Moving parts are the main problem

[ThePhilips]

The article you cite does not contain the 1-1.5 years figure anywhere.

Like all articles on Anand, it's a performance benchmark. Obviously there is no reliability numbers in the article.

How did you get that number?

Honestly, I have picked it up from discussions with other folks who have been using SSD for quite some time and actually pressed one tech support employee to provide a number. Apparently EOMs do not have numbers themselves (or no numbers they can reveal to customers), so the support guy gave his wild guess of "1-1.5 years". Doubt the number all you want. We can only hope for the drives to hold that long.

I have tried to calculate something by myself just to find that I'm no good with the numbers myself.

Let's take 64GB SSD with 100% working wear leveling. Presume that we write ~32GB of data per day (realistic, as small files updates are the main load in e.g. desktop environments). (XXX) That means that every two days whole SSD is written and the infamous "100K erase ops" counter decreases. That is 366/2 = 183 decrements per year. Peanuts compared to 100K of erase cycles. Even if we are to factor in wear level algorithm with realistic efficiency of 90-97% (as per Anand research) and that , it still results in too long longevity.

But now at the "(XXX)" point we obviously made a mistake. SSD isn't going to be overwritten completely every two days. E.g. installed OS and user files are already taking place. That means that wear leveling can only work on the free space - but not whole SSD capacity. IOW, if there is only 5% of SSD space is available, then the 5% are going to be constantly erased/written - probably much more often than once per day.

In the end, picture is even more murkier than it was in the beginning.

Re:Moving parts are the main problem

[SlowGenius]

I don't know how it's currently done, but I can think of a trivial solution to that much of the problem, at least: Not only do you keep an index to the current location of block 0, but you keep a counter with it. And maybe an index and a counter to that index. And maybe an index and a counter to that index (etc.) "But wait!," you say, "you *still* have to update the counters and the indices to the indices, and, and..." Yes, you do. But you can do so in a logarithmic fashion, so the higher up the chain of indices a counter/index pair is, the more static they are. With this approach, ultimately your "root index" can change as infrequently as you want-- averaging out to once a year (or even less) if you so desire.

Re:Moving parts are the main problem

[Retric]

Not bad but a better solution, use an Index that points to the location of the start of the index.

Now you only need to change that index after the old one get's used up in ~50k writes. So now you have 50k * 50k = 2.5 million write cycles on the index. If that's not enough use a 3 stage index and you have 1.25 trillion writes to the index.

Re:Moving parts are the main problem

[Zashi]

Quit using windex/bleach on your computer mother boards!

Re:Moving parts are the main problem

[sybreon]

I second the virtual machine idea. That would give you the most guarantee of running the system through the years. Regardless of how hardware interfaces may change, you can always virtualise things. I had to do something similar recently - moving an older application onto newer hardware. KVM really saved my skin.

Re:Moving parts are the main problem

[cenc]

I would throw in go for something like HTML with XML or text data storage, and then stick it on multiple machines. Planning for 15 years is noble, but planning for something that can be easily converted and moved to a new system is more realistic. The more you can distribute it, the more likly it is to keep on functioning.

If you have to have it bound to hardware, and you have two computers to work with, buy two computers that are exactly the same. Say in 10 years, you can combine the parts when one dies. In the meantime spread the storage over the two machines.

Re:Moving parts are the main problem

[jimthehorsegod]

Um... Why'd you need anything like that? Apache, locally installed on one PC designated as a 'server'?

Re:Moving parts are the main problem

[mysidia]

By my definition, the most reliable storage setup you'd see on a desktop is a hardware RAID1 array with 2 Ultra160 or Ultra320 SCSI-based SSD drives.

And the best cooling setup is a system that doesn't require cooling fans, due to its setup, but has ample cooling measures, just in case.

The underlying assumption here is that the highly reliable hardware is also the more expensive equipment. Although not everything that is more expensive is also more reliable.

That's because engineering and ensuring the highest level of quality is more expensive to do.

In terms of computer equipment, the most reliable equipment is that designed and intended to be used in servers that run with heavy loads 24 hours a day, 7 days a week.

The highest reliability comes when the system is designed so that some parts can fail; i.e. Redundant Hard drives, Reduntant RAM, Redundant CPUs, Redundant Power, good power protection.

Despite all this, the mainboard can still fail. But the more carefully designed more carefully tested parts are less likely to do so in the long run (assuming they survive the first 90 days)

Re:Moving parts are the main problem

[KillerBob]

You're right. You're not expecting a drive to fail every 6 weeks on the button. But MTBF is used for logistic planning... in the given example, you *should* be ok having 2 spare drives on hand. One to replace the one that you expect to die, and one to replace the one that dies while you're waiting on the replacement you just ordered to arrive.

There'll always be statistical skew when working with something like a mean time, and it is an oversimplification. I've only ever had 2 hard drives fail on me, for example. But that's the point with MTBF... to give you an idea of what kind of fail rate to expect so that you can plan how many spares you need to keep on hand. It's got nothing to do with the actual expected lifespan of an individual drive.

Re:Moving parts are the main problem

[Larryish]

Is virtualization a possibility?

Why not go to 3 machines, 2 of which run vmware client and one which runs vmware server?

Then the hardware is unimportant.

Re:Moving parts are the main problem

[eudaemon]

I have tried OpenSolaris w/ZFS on a 8GB flash drive. It was incredibly slow. Incredibly, incredibly slow.

Re:Moving parts are the main problem

[commodore64_love]

>>>I'd say an online UPS as a component to help prevent premature power supply failure.

IMHO ye are over-engineering this whole project. Instead I'd recommend the approach I used six years ago: (1) Buy the best product you can with the latest CPU (i7). (2) Max it out with as much RAM as you can afford so there's room for future software bloat...er, growth. (3) Buy a good product, not an eMachine.

The end. I've had my PC for seven years now and although it's started to feel a little cramped (1 gig/3000 megahertz), it still runs anything I want it to run, and I fully expect it to be working in 2012. It appears this dentist followed the same philosophy when he originally bought his machine in circa 1993.

Re:Moving parts are the main problem

[The_Wilschon]

Regardless of the failure rate of SSDs, this veterinary office took 15 years to fill up 500 MB! They almost certainly aren't doing enough writes to wear out an SSD before the cows come home.

And yes, I know you can do an awful lot of rewrites without filling up any extra space (an unbounded number, in fact). But just spend a second thinking about the likely workflow of a veterinary office. I still don't think they really will be running a large number of rewrite cycles.

Re:Moving parts are the main problem

[Anonymous Coward]

Get well-designed fans? Might not worth the trouble for computers, but we get them for self-built scientific equipment with potentially long lab life.

Actually, these fans you mention do not have a MTBF that is high enough for a 10 year lifetime application.

They say their MTBF is greater than 150000h.

A rule of thumb for failure propabilities is:
L1 : L10 : MTBF = 0.21 : 1.0 : 3.0

So a batch of them has an L10 (10% chance to fail) of 50000h. Which roughly equals 5.7 years. Nowhere do they say at which temperature they measured this MTBF, which is very important as failure is more likely, the hotter the fan runs. We will have to assume they test at somewhere about room temperature.

These fans are a better choice in that respect, they have an L10 of 80000h operating at 40 degrees celsius, closely resembling the intended product lifetime at operating conditions.

At least in Germany, both have roughly equal prices (I think it's somewhere about 20 euros).

These are a good choice for long lifetime products, too. Prices are similar to the above mentioned products.

Re:Moving parts are the main problem

[Anonymous Coward]

I just bought a 32gb Samsung SSD for $120 bux.(you can get a 8G for $75 or so) Somehow I think when replacing two computers, this is hardly a major expense.

Regarding failures and failure rates. Raids, SSD, whatever, is NO SUBSTITUTE FOR BACKUPS!

Raid or SSD or whatever will not save you from human error, will not save you from major hardware failure, will not save you from your house or business burning down, ect...

Last thing. How much is your data worth? If you are going to make arbitrary formula's with prices of multiple disks versus prices of SSD, shouldn't the value of your data be in this equation? IMHO data value generally exceeds value of storage device by several times.

Re:Moving parts are the main problem

[tom17]

Block rotation in the 21st century was CONSIDERABLY more widespread after...

Re:Moving parts are the main problem

[Anonymous Coward]

"... so the flakiness goes away. Install all updates..."

Where would you get updates for WIN 95? Back up the software twice, convert one backup to a commonly-used database (if it isn't already), store copies of that on disk and a web-based service, and wait for the old box to die. When it does, replace it with a new unit, loaded with a currently produced & supported OS.

Re:Moving parts are the main problem

[cthulhu11]

There's pretty much no way to know whether a given device has an "up to date" controller.

Re:Moving parts are the main problem

[PingPongBoy]

So actually the whole problem is an industry problem. The desired number of years (15) is somewhat arbitrary. Why not 100 years? Plants are designed to withstand storms and events that take place once per 100 years, though minor items might fail from time to time.

Industry looks at the needs of people: people think of replacing their computers every few years, but might not actually do it so the computer industry makes things that don't all last for decades. Things like power cords and keyboards might take ridiculous amounts of abuse because they're relatively large compared to Planck's constant but Moore's law makes it far more attractive to replace the high-end parts.

The vet is looking for a high-end cash register (I do know more about vet computing because I wrote software for animal diagnostics-but vets basically have a database of test results and bill owners for procedures and tests; there's little in the way of dealing with pharmacies, health insurance, etc, etc). Not so many decades ago, cash registers, even the electronic ones, don't appear to die at a moment's notice. They take a pounding from people opening and slamming the drawers, as well as wedging rolls of paper into the innards, and steady maintenance was required to keep everything in shape. Today, the cost of maintenance is replaced by the purchase price of a replacement. Cash registers used to cost thousands. Computers used to cost even more. Maintenance was at least 50 bucks per visit. If computers only cost about two hundred for a new one capable of running DOS, and come with three-year warranties, the economics say that the computer might be expected to survive for twice the warranty period-why go to the extra trouble of designing an over-the-top machine that lasts just twice as long (12 years, stretchable to 15) as the most basic machine you can find? As the machine gets older, the maintenance becomes even more oppressive and the total cost of ownership would be in the thousands. About 10 years from now, computer interfaces would probably be available anywhere in the form of holograms talking to a server anyways. Who's going to run a dedicated machine as a cash register?

Re:Moving parts are the main problem

[maxume]

Buy it from a reputable dealer and make sure it says "Intel" on the box.

AnandTech also seems to have a pretty good handle on things.

Re:Moving parts are the main problem

[duplo1]

You're right about the moving parts, however I disagree about the flash memory bit. Electrical utilities deploy systems to substations, which utilize flash-based storage systems and they are sold with a 10-year warranty. Having seen devices with rotating storage fail in those environments when flash-based systems have not, I think that it's fair to say that flash memory will definitely go a long way in the reliability department.

Re:Moving parts are the main problem

[Sj0]

I'm the head of the Reliability Centred Maintenance program at the industrial plant I work at. In RCM, we look at the dominant failure modes, and prescribe a maintenance program to mitigate the risk, or reduce the frequency.

In this case, the "I want this computer to last for 15 years" implicitly means they don't want to do scheduled maintenance. They want it to sit there and run, like the previous machine. They don't want a PC in the way you or I think of a PC, they want an appliance that just works. That being the case, We NEED to look at reliability centred design, rather than maintenance.

So what are the dominant failure modes for a PC? Clogged fans, failed power supply, hard disk failure. If you don't experience these failures, odds are your computer will run indefinitely.

The first problem can be solved with a machine that doesn't have any fans. Design your machine so convection currents carry the heat out the top of the case. This will mean you'll never have a fan failure.

The second problem can be solved with two methods: First, redundant(fanless) power supplies. Second, an online UPS to prevent dirty power from damaging the machine. I might actually just use an industrial deep cycle 12V battery with a pair of inverters, and a 12V smart battery charger on the AC side. It's dirty, but it's functional. Your charger should last 15 years, your battery should last 20, your inverters should last indefinitely and are redundant. With these two solutions in place, I wouldn't expect a total system failure for 25 years. If the charger fails, you should have more than enough time running a 50W fanless PC and 50W lcd monitor to schedule replacement of the charger.

That leaves the hard drive as the only remaining failure mode. Hard drives aren't going to last 15 years. I had a hard drive from 1989 that lived to see the new millennium, but it's dead today. Along the way, many of its contemporaries decided to die. The only solution is to mitigate the consequences of failure with redundancy, so the drive can be replaced. A CompactFlash drive might be a good option, but the standard itself is only 15 years old today, so it's difficult to say whether such a solution would work. With this solution, you would probably need to replace a drive every 7 years, but it could be done during a scheduled outage, outside of office hours.

If you're serious about reliability, leaving it to luck is a good way to be negatively surprised. I've worked with too many failed PCs in the past few weeks to believe you can just build it and forget it.

Re:Moving parts are the main problem

[Geste]

"Assuming that you have to rewrite his software, make it all web based"

Making things Web-based is a great general goal but is it worth it for a 2-node business system? Not likely.

And does it really need to be rewritten? If it is doing the job now, I'd look a little harder at the existing DOS spplication. If you discover that is some old dBaseIII-Plus (or Foxpro or Clipper) application, you might be able to just transplant it onto a Linux box with something like Flagship or Harbour.

Re:Moving parts are the main problem

[Anonymous Coward]

Please. "an SSD" not "a SSD drive".

Re:Moving parts are the main problem

[po134]

on the hardrive I suggest buying RAID-edition or other ENTERPRISE-class HDs as they went throug more testing (like the serveur cpu class such as opteron or xeon).The best thing would be to have raid-0 installed. (most mobo have this option these day) CPU and motherboard are the one that are the hardest to replace in my opinion.

Re:Moving parts are the main problem

[GregNorc]

Top of the line will generate a lot of heat though. OP doesn't need lots of processing power... an Atom processor would be better.

This would reduce the chances of heat related failures... shell out for good fans and a good PSU. A RAID 1 would be sufficient... you don't sound like you're dealing with a lot of data, it might be better to show him how to use a DVD-R burner and have him manually back up every month or so. If you do go for the RAID, stick to tried and true technology (No bleeding edge 2TB drives. Also sticking to say, a 5400rpm drive will reduce chances of drive failure.)

Re:Moving parts are the main problem

[Lennie]

If using Linux, mounting with noatime could also help prevent writes to disk.

Re:Moving parts are the main problem

[jonbryce]

Basically, anything new is likely to be less reliable than something that is tried and tested.

Eg, Windows XP is more reliable than Vista, and much more reliable than it was when it first came out.

Re:Moving parts are the main problem

[mr_mischief]

A cheap single-channel SCSI card and a DDS3 or DDS4 SCSI tape drive should back up any drive you have in a Pentium 100 with plenty of room to spare. Alternatively, an IDE CD-RW drive is about $15.

You might not have a way to add an IDE port to the 486 if you are out of them or to add the SCSI card, but there are plenty of ISA, EISA, PCI, and even VESA-LB cards for either on the net that are working pulls or surplus factory-sealed stock.

Check out WeirdStuff, PC Surplus Online, or PacificGeek.com for what you need. Just please don't buy the last of the ISA SCSI cards before I get another one. ;-)

Re:Moving parts are the main problem

[endymion.nz]

Same here. I've had a PC of some description since 1986 and I've had one mainboard failure, one HDD failure, one 3Dfx Voodoo texture memory failure, and that's it. Most of my PC's over the years have lived into obsolesence and been sold (or in the case of the last one, swapped for a Wii :D)

Re:Moving parts are the main problem

[Annirak]

It'd sure be nice if you knew what you were talking about.

Power supplies do not fail randomly. They fail as a function of wear and abuse. Heat is the primary determining factor, as with most electronics.

Raid has no beneficial effect on SSDs. The failure mode of an SSD is FAR different from that of an HDD. Because SSDs fail to a read-only mode, there is little point in running SSD RAID for data security.

SSD's DO NOT use battery backed RAM. They use capacitor backed RAM. (I've opened one up, and there were two 4.7F Aerogel capacitors, no batteries). This drastically changes your battery failure claim.

Your concept of "reasonable size" and the OP's are different. Remember, his original drive which has been in operation for *15 years* is only just now filling up. Potentially, a 1GB drive would last another 15 years. 32GB will be completely suitable and "reasonable."

Rewrite can be far less of an issue than you think. Intel's ssd (admittedly, this is the most expensive one on the market) is rated for 100GB/day of continuous writes for 5 years. XP or Vista will work just fine in this environment. Much like a previous comment, I would suggest underclocking a processor, which makes this machine fundamentally incompatible with Vista. XP or Linux would both be good options. DOS emulation on both platforms is well established.

Switching gears now, to the OP:

For a processor, you should stick to media-center processors. These processors are designed to be very low power in order to keep the noise level from fans down. If you couple a low power processor with a passive cooling solution such as thermalright cooling towers and a 120mm fan duct, you can keep the processor fan out of the equation.

Avoid moving parts wherever possible. There are a few power supplies available that use heatpipes to bring the heat out of the supply, to a radiator grid that protrudes from the back. There are not, unfortunately, equivalent solutions for the CPU fan. You also need air circulation over your motherboard. This means that you need to be very careful about what kind of case you buy. The case needs to have dust management because dust is the primary cause of failure in fans. Try to avoid cheap sleeve bearing fans, and buy either ball bearing or hydro-wave bearing fans.

To summarize:
Processor: low power, underclocked/undervolted processor, AMD used to have some 35W processors, but those should be available from intel as well now.
Heatsink: Use a heatpipe cooler with a duct to the rear exhaust fan
Motherboard: No electrolytic caps, use IGP only if there is a heatpiped cooling solution.
RAM: use heatsinked RAM, underclock & undervolt if possible.
HDD: Use an SSD with a big write tolerance (intel's 100GB/day is good). Consider creating a ramdisk for intensive activities, get the OS to store the ramdisk to the drive on shutdown, and restore it on startup.
Power Supply: Try for a passive cooling solution, but do not go over a factor of 5 above your average load (e.g. your computer draws 100W, do not go over 500W). Underloading supplies can lead to early failures.
OS: You're stuck with an emulated solution regardless, since this is a DOS application. Might as well go with Linux, since you have better control over reads/writes.

Here's one I haven't seen mentioned yet:
Put it on a UPS and configure it to initiate immediate, safe, shutdown on power loss. This will cut down on failures due to power loss.

Re:Moving parts are the main problem

[bandmassa]

Maybe an alternative way is to image the disk of the old 486 and run it in a virtual machine, backing up the image and VM regularly, then it doesn't matter what hardware you run it on, and the hardware can be easily and quickly replaced if it does fail. 15 years out of a workstation probably requires an SSD instead of a hard disk and wind the performance settings way back so the fans hardly ever run. It certainly would require good build quality, and it'd be a struggle to find that these days in even a name brand like Dell Lenovo or even Apple.

Re:Moving parts are the main problem

[Anonymous Coward]

So how many write cycles does that index data have?

A lot. Which is why you use interesting tricks for storing the index too. Google "log structured file system" to find examples of filesystem designs which don't overwrite old index entries, and instead append new information to a log. Good SSDs use something similar to structure data on disk.

One thing to keep in mind while thinking about the possibilities for the algorithms is that NAND flash chips of the type used in SSDs have some extra scratchpad storage for each block. Its nominal purpose is to store ECC data since high density NAND (especially MLC NAND) has a substantial error rate. However, there's enough of it that the controller can also store a little bit of metadata associated with that block. Obvious uses of this would include ID tags (indicating if the block is index metadata or user data), logical (host-visible) block numbers for user data, pointers, and flags.

One more thing which is relevant: the nature of the operations which can change the state of a flash bit. There are two, erase and program. Erase takes a very large block of bits and resets them all to logic '1'. Programming is done on a smaller scale, and when the hardware actually performs the write, it really only does things to the bits which are written with a logic '0'. Writing a '1' means it doesn't do anything to that bit, since there's no way for the program command to change a bit from '0' to '1'.

Say you have a freshly erased byte: 0b11111111. The implication of the above is that you can send many write commands for that byte, programming individual bits at a time, and as long as you don't program any given bit more than once, the byte as a whole experiences no more wear than programming it with all zeroes in one operation. If you first write 0b11101111 to that freshly erased byte, then write 0b11111110, the byte will contain 0b11101110.

You can use this to support log structured storage. You'd erase a block and incrementally write small sub-block-size log structures to it. Whenever you supersede a log entry with a newer one, you can program a flag in the old version to indicate it's no longer valid. You could even have a reserved pointer field in each log entry, to be written with the physical address of the new log entry which supersedes it.

This does mean that SSDs often need to scan through complex data structures in order to figure out just where the current valid copy of a given logical block resides. However, the overhead for this is low, and most SSDs have at least some DRAM attached to the controller to cache metadata.

The SSD controller's code isn't easy software. It's tricky and difficult to get really right. For example, I didn't address the problem of reclaiming space consumed by dead log entries, which turns out to be a huge issue. There are lots of "bad" SSDs out there: drives which suffer from the so-called 'stuttering' effect. Writes to such drives often cause the drive to stop responding for a very long time (some of them have been measured at 1 second or more worst-case write latency). Just like garbage collection, if free space reclamation isn't designed very carefully, every so often the drive is going to have to spend a relative eternity on housekeeping before it can actually complete a write.

AnandTech has been running some good reviews highlighting which SSDs actually address this issue correctly. Intel's X25-M drives are the only MLC SSDs which can currently sustain high random write IOPS (I/Os Per Second). There are some others which get it right enough to avoid stuttering and still outperform conventional HDs in random write IOPS, though only by a factor of 2 or so. And then there's the run of the mill SSDs with JMicron controllers which suffer from stuttering and whose performance drops way below a HD's on any load which involves even a little bit of random writing.

Re:Moving parts are the main problem

[ShakaUVM]

>>By my definition, the most reliable storage setup you'd see on a desktop is a hardware RAID1 array with 2 Ultra160 or Ultra320 SCSI-based SSD drives.

Sure. I use RAID1 on all my business machines. The extra hard drive isn't expensive, and it guards against HD crashes. What it doesn't guard against is mobo crashes, and if you lose your motherboard, all the data on your RAID array vanishes unless you can find a replacement with an identical chipset. Good luck with that when you're building a machine to last for more than 5 years.

So I actually use a RAID1 array and an external hard drive with nightly backups.

>>And the best cooling setup is a system that doesn't require cooling fans, due to its setup, but has ample cooling measures, just in case.

I'd disagree with this. Passive systems often run much hotter than actively cooled systems, and thus increase your failure rate. Depends on your build, of course, but it's what I've seen from experimenting with trying to build one for silence reasons. Even underclocked, I just couldn't get it as cool as with fans pushing the air out.

>>The underlying assumption here is that the highly reliable hardware is also the more expensive equipment. Although not everything that is more expensive is also more reliable.
>>That's because engineering and ensuring the highest level of quality is more expensive to do.

Yeah, but cheaper parts are often the most successful, with the economy of scale pushing the prices down. I don't think there's much difference in reliability between Intel's most expensive and least expensive offerings on the market right now (though I could be wrong). Perhaps if you're looking at el-cheapo DVD drives or something. From some EEs I worked with, there were indeed major quality issues with some of the knockoff brands (they'd reverse engineer whatever Sony was doing, but never really get it right). So in that case, I guess sticking with a Sony drive or whatever would be your best bet.

>>In terms of computer equipment, the most reliable equipment is that designed and intended to be used in servers that run with heavy loads 24 hours a day, 7 days a week.

Again, I'm not sure about that. Enterprise hardware, because it looks at less customers than consumer products, often have bugs that go unnoticed. A friend of mine works for Hotmail, and they found a number of firmware bugs in various enterprise level products they used, including a very expensive RAID controller capable of running a lot of disks in parallel - it would cause a complete data loss with about a 50/50 probability once per year.

>>The highest reliability comes when the system is designed so that some parts can fail; i.e. Redundant Hard drives, Reduntant RAM, Redundant CPUs, Redundant Power, good power protection.

If you have a dual-CPU motherboard and one of the CPUs or RAM modules fails, I believe most mobos will refuse to boot. In any event, when you have doubled everything you've greatly increased your risk of failure. Remember, the OP wanted a machine that could run without maintenance for 15 years - having to go in and pull bad components defeats the entire purpose.

>>Despite all this, the mainboard can still fail. But the more carefully designed more carefully tested parts are less likely to do so in the long run (assuming they survive the first 90 days)

Yeah, I've had more mobos fail on me than any other component... small sample size, but it sort of reinforces my point about not knowing what is reliable - all the mobos made with bad caps a couple years back certainly seemed reliable under testing, but then would suddenly fail a couple years later when they'd blow a cap. As a customer, you have no way of knowing if any of the current products on the shelf at Fry's have similar ticking time bombs.

Re:Moving parts are the main problem

[mysidia]

Power supplies do not fail randomly. They fail as a function of wear and abuse. Heat is the primary determining factor, as with most electronics.

Simply put, wrong. Wear and abuse are essentially constants for a given system design. The duration before eventual component failure is still random, with certain probabilities of different durations, depending on the specific power supply and system properties.

You seem to be forgetting that all components have an approximate expected lifetime and a random failure component.

Every electronic part in the PSU is subject to random failure, in addition to the possibility of abuse-induced failure.

Because SSDs fail to a read-only mode, there is little point in running SSD RAID for data security.

You're also wrong here. SSDs can completely fail much like RAM can; statistically it will probably take a long time before most SSD arrays experience catastrophic component failure, but it's an unwarranted risk. RAID is beneficial b/c of the importance of the data.

SSD's DO NOT use battery backed RAM. They use capacitor backed RAM.

Capacitors are the same technology as batteries and have a limited lifetime as well. Not all SSDs are created the same; some use capacitors, many use other types of batteries.

Potentially, a 1GB drive would last another 15 years. 32GB will be completely suitable and "reasonable."

An unwarranted assumption since he is changing platforms, partly due to space constraints. If you only have 500mbs, then you go to great lengths to conserve space. Once you're running a modern OS (Windows Vista, Office 2007), and saving lots of documents, 30gbs doesn't seem such a guaranteed thing anymore.

Rewrite can be far less of an issue than you think. Intel's ssd (admittedly, this is the most expensive one on the market) is rated for 100GB/day of continuous writes for 5 years.

This rating is very likely far overoptimistic. Not everyone's write patterns will be the same.

Avoiding moving parts is fine, provided you don't wind up with something that's hotter.

Moving parts fail, but they're generally cheap to replace, unless on a power supply.

Use CPU and Case fan / mobo combos where the mainboard can monitor CPU and chassis fans and force a shutdown if the fan fails.

If this keeps the system cooler, then the whole thing will last longer, even if a fan needs to be replaced some day.

Re:Moving parts are the main problem

[unitron]

A lot of that dust probably came out of the vacuum's exhaust. If air molecules can be rammed through the filters, so can little bitty dust particles.

If you use a vacuum cleaner, you don't have to go outside. A soft, natural bristle brush is also helpful in urging the dust into the hose.

Re:Moving parts are the main problem

[unitron]

So whether moving air creates static electricity depends on which direction it moves? Regardless of relative motherboard orientation?

Use a soft, natural bristle brush to loosen the stuff and a vacuum to remove it.

Re:Moving parts are the main problem

[supernova_hq]

I would love to see the little bristle brush that you have that will get dust out of MY headsink. That suckers about 3.5" deep! And no, puting the vacuum (on suction) directly on it does NOT get all the dust out.

Re:Moving parts are the main problem

[Anonymous Coward]

Just open up the old computer, dust it out. Back up the drive while your at it. Then put it back together. If you have the room in the case and the plugs, piggyback another drive (a 30GB should last a good while space-wise when replacing two 500MB drives on a Win95-Win98 computer) or two because running out of room for files is likely the only real problem. The old computer is likely to keep truckin' on.

I've seen some old-assed stand-alone computers at places like car repair shops or small businesses, still running Win98 SE (basically what amounts to Win95 SP3 with less flakiness.) Typically if they last that long and are kept away from the internet (because malware can cripple a legacy machine), they don't die unless somebody takes a hammer to it or lightning strikes.

If you need something newer and faster, go with a midrange setup and underclock to keep it cool. The high end will be way too fancy and expensive for what's needed, and low end is likely to skimp out on the quality of some parts. Just back up regularly, dust every now and then, and perhaps having a UPS to prevent unexpected shutdowns will do a lot to keep a halfway decent computer going for a loooong time. Considering the last computer lasted 15 years, I figure somebody already has a clue in that department.

Re:Moving parts are the main problem

[unitron]

House painting brush, perhaps augmented with pipe cleaners or Q-tips, or very skinny artist's brush.

Re:Moving parts are the main problem

[Anonymous Coward]

I fix PCs in various locations and the dirtiest I have -ever- seen was one in a vet's. Absolutely full of dust and fluff -no wonder it had slowed to a crawl. A major requirement is an air filter - one of the cheapo air conditioning filters is all that's required, just vacuum it off every so often. A passively cooled system works best with a chimney to promote convection current.

Re:Moving parts are the main problem

[cbiltcliffe]

So whether moving air creates static electricity depends on which direction it moves? Regardless of relative motherboard orientation?

No. The flexible plastic hose of most vacuum cleaners generates static as air and heat-dried dust travels through it.

Use a soft, natural bristle brush to loosen the stuff and a vacuum to remove it.

That's the only way I'd ever use a vacuum on a computer. And even then, it's iffy. The bristles of the brush can drive dust in further just as effectively, possibly even more so, than an air compressor.

Re:Moving parts are the main problem

[cbiltcliffe]

House painting brush, perhaps augmented with pipe cleaners or Q-tips, or very skinny artist's brush.

Or just save all the freaking trouble, and use an air compressor.

Re:Moving parts are the main problem

[cbiltcliffe]

If he is using Win95, M$ stopped supporting that years ago; I'm pretty sure there are no Windows updates for that OS on their site anymore.

Haven't looked around their site much, have you? There are still updates for Windows 3.1, so I think the 65 or so updates for Windows 95 will be around for a bit.

And even if they're not, I have them all downloaded....

No...the updates are definitely there. There just aren't any new ones.

Re:Moving parts are the main problem

[cbiltcliffe]

Where would you get updates for WIN 95?

Errrmm.... microsoft.com? All the Win95 updates are still there, and still available through windows update....

Re:Moving parts are the main problem

[unitron]

And then wonder why the next PCI card or DIMM you plug in to a previously empty socket isn't working right.

Re:Moving parts are the main problem

[supernova_hq]

I have used compressed air on many VERY dirty computers and NEVER had that problem. The only holes I can think of on a MOBO (or in the machine for that matter), that have contact areas small enough to be blocked by such small pieces of dust are the CPU SOCKET. And if dust is getting in there, you are missing a very integral part of your computing experience...

Re:Moving parts are the main problem

[Kadin2048]

> Tapes are slow, unreliable and expensive. I would not use them for any purpose.

I'll give you the slow and expensive part, but unreliable? That I'm not sure I'd agree with at all.

With the possible exception of helical-scan DAT, which was a good idea poorly implemented in many cases, tape drives and data stored on tape is extremely reliable. You can record data to a magnetic tape, put it on a shelf, and it will still be readable in 20 years. Heck, it'll probably still be readable in 50 or 100 years; the limiting factor is really the drive, not the tape. And enterprise-grade tape drives (practically all that's being manufactured anymore; they've pretty much cut it out with the cheap DATs) are built quite well.

Hard-drive based backups are very nice and I do think they're appropriate for most users, but tapes still have a place. If you need to store a lot of data either offline or nearline, the operating cost of a tape library is very low. (They don't require any electricity just sitting there, and you don't have to worry about periodically spinning them up or anything.) For archival purposes they are also excellent.

I'd perhaps even concede that tapes are a poor "backup" media, but an excellent "archival" one, if you were to split those functions out separately.

Where I would never use tape is for nightly backups that are just going to be overwritten the next day, but for saving data for the long term they are a good low-maintenance option.

Re:Moving parts are the main problem

[Kadin2048]

Question regarding the reliability of fanless computers: wouldn't a fanless system eventually overheat and die if it wasn't dusted periodically?

It seems like 10 or 15 years of dust in a typical office environment, even without fans moving a lot of air around, could seriously impair the effectiveness of heatsinks and cause failure.

If you went this route and really wanted zero-maintenance, I'd think you'd need to factor in dust and over-engineer the heatsinks significantly, so that you'd have the radiative capacity to stay below temperature cutoffs for the full lifetime of the machine.

I wonder if somewhere, somebody has tables showing dust buildup and its effects on electrical equipment over time in typical office environments...seems like it must be a well-thought-over issue.

Re:Moving parts are the main problem

[mdwebster]

First spec sheet I looked up from STMicro stated 1 million write cycles and 40 year retention. Second one from Microchip says a million as well. Third one from Atmel says a million cycles + 100 years retention.

But it's worth thinking about for a while to come up with a flash solution since even adding a $.10 IC adds up over millions of drives.

Re:Moving parts are the main problem

[mcmonkey]

In this case, the "I want this computer to last for 15 years" implicitly means they don't want to do scheduled maintenance. They want it to sit there and run, like the previous machine. They don't want a PC in the way you or I think of a PC, they want an appliance that just works. That being the case, We NEED to look at reliability centred design, rather than maintenance.

But at what cost? For example, if I wanted to buy a car to last 15 years of regular use, I could find many suitable candidates. Many cars on the market, available at a reasonable price, will last 15 years with proper regular maintenance. Even accounting for the cost of maintenance over 15 years, this is a reasonable request. On the other hand, if I wanted a car to last 15 years of regular use with no maintenance other than fuel, that is a much different request. Brakes that last 15 years, tires that last 15 years, belts and hoses that last 15 years--the cost of these things will be exponentially higher than buying regular off the shelf parts and replacing them over time. Even if one time I had that one car that went 15 years with no maintenance, it would be unreasonably expensive to meet that requirement. We're not talking about NASA. We're not building computers for nuclear subs or deep space probes. Maybe this guy did buy a computer in 1994 and it lasted 15 years. But that does not make it reasonable to expect every computer to be as reliable. The golfer who gets a hole in one on her first shot shouldn't expect to get a hole in one on every shot. This guy shouldn't expect every computer to last 15 years without maintenance.

Buy any current workstation and...

[Anonymous Coward]

Virtualize!

Then your father's old setup can remain DOS and Win95 effectively forever, on any modern hardware. I've done this for lots of clients with legacy WinNT and Win95 systems.

The process is called "physical to virtual" (P2V) migration.

Re:Buy any current workstation and...

[amcdiarmid]

Mod parent up.

You need DOS and Win96 compatibility: You can virtualize the existing system into a new system, and make it portible; back-up-able (as a Virtual Machine) by virtualizing it.

As an aside, I always thought Win95 was a dog. You may wish to check to see if XP compatibility mode will work, or check (ha ha) to see if WINE will work. (Actually, trying the application set with WINE is not a bad idea - it should be compatible with Windows 95 by now.)

Remember it could be worse: I have a friend who deals with Vet who has an old Xenix system - they buy parts of ebay in bulk;)

Re:Buy any current workstation and...

[yerktoader]

I would guess that it would be a better idea to do both a VM AND build a solid machine with backup parts. It's not guaranteed that a given VM developer will be around in 15 years, nor is it guaranteed that hardware sitting unused in a environmentally controlled room will fire up in 10(though I would expect it more than likely to be just fine).

I would try to stick to parts that are verified by the motherboard manufacturer as compatible. As others have already stated, components with moving parts are likely the first to fail so those should be high on the list of backup parts. Power supplies are another already mentioned, though I suspect most failures among high end power supplies are either due to heavy use such as gaming, and lack of maintenance - dust and moisture really are killers.

I'm wondering what the outcome of the AMD/ATI - Intel - NVidia war is going to be. In 10 years you could end up looking at a market of PC's running on NVidia processors or a whole new dynamic altogether - if Ray Kurzweil is right, though that's pure speculation.

I figure if you both virtualize the former server on new, decent hardware with a few spare parts - probably at least one to two spares for each component - with proper maintenance you'll more than likely be just fine.

Re:Buy any current workstation and...

[Anonymous Coward]

Xenix? Yuck. He's probably running old AVS Better Choice. I used to work for them and support that pile of shit. It's written in COBOL! Nasty.

Re:Buy any current workstation and...

[m509272]

Absolutely. Great idea. Set it up, have VM copies on multiple HDs and solid state that you test periodically. Make sure that his application VM isn't contaminated by loading "new" apps. Periodically switch to "new" hardware as need be. I might suggest a Linux OS as host OS.

Re:Buy any current workstation and...

[LWATCDR]

Sounds good to me. DosBox or VirtualBox running under Linux would work just fine depending on if you want to run it under Windows 95/98 or Dos.
With Virtual Box you could even do a complete backup on a USB key and maybe even to the "cloud".
I wounder if you could even run VirtualBox on EC2 and us S3 for making backups.

Re:Buy any current workstation and...

[linhares]

You can also use dropbox to backup a virtualbox machine to the web and keep it in synch with other machines.Works beautifully.

Re:Buy any current workstation and...

[linhares]

And you'll get even more space if you use a referral code like this.

Re:Buy any current workstation and...

[Anonymous Coward]

I was going to say something close. This is definitely the solution I'd choose. It allows for the most flexibility in keeping the legacy system and ability to upgrade the entire system if he so desires.

Re:Buy any current workstation and...

[linhares]

Yes, you can, through dropbox.

Re:Buy any current workstation and...

[FishWithAHammer]

Stop spamming your referrals, asshat.

Re:Buy any current workstation and...

[Anonymous Coward]

Sounds like a good idea - but not necessarily just for preserving the Win95/DOS systems. Feel free to update to a recently released OS so you're still in support, like Win7 which should be out sometime soon, but run the systems in a virtual machine. Then, backup the VHDs nightly onto a third machine or NAS. Separate your hardware from the software and don't worry about your workstations. You can just buy commodity hardware at any point in the future if there are issues with the initial systems and copy over the VHDs and you should be back and running.

Re:Buy any current workstation and...

[Eil]

Not knowing much about the technical details, I was going to suggest any old Linux distribution and then running DOSEMU + FreeDOS on it, but virtualization probably is the far superior option, for the following reasons:

You can continue using the exact same setup, only running on virtual hardware.

Backups are a brain-dead simple. Just shut down the virtual machine and copy the image. You can even save snapshots and such.

Restores are likewise trivial. A virtualized OS "sees" the same hardware no matter where its running.

The native OS can be kept up to date with all the newfangled web browsers and such without worrying about how it'll affect the stability of the database application.

To be honest, I would consider upgrading the virtualized OS to Win98SE at least. If only DOS is required, perhaps consider tossing out Windows anyway and using a late version of MS-DOS, OpenDOS/DR-DOS, or FreeDOS. (All have their pros and cons but if this is a fairly standard DOS program we're talking about, it should run fine on any of those.)

To the submitter: This actually sounds like a fun project. Good luck.

Re:Buy any current workstation and...

[Nimey]

I once worked on a veterinarian's Xenix system. Came to us with a hard drive problem, IIRC. The fun part was that whomever installed Xenix decided that they didn't need man pages or any sort of documentation. Since my only *nix experience up to that point was Linux, it was... interesting.

Re:Buy any current workstation and...

[linhares]

Oh well, sorry. I did it with good intentions, but it's spam, I guess. I was thinking: provided someone actually joins dropbox, they would be actually getting more space. Oh well, I guess I deserve the troll mod; but if anyone joins it and uses it, you know what? Then I just might have done someone a favor. This is not zero-sum. And I do believe that the best solution here is virtualization with a dropbox synch... so despite the spam, for which I apologize, that's really my opinion.

Re:Buy any current workstation and...

[molarmass192]

Imagine what the tech landscape would be like if MS had stuck with Xenix ... what a totally bizarro world moment!

Re:Buy any current workstation and...

[wesborgmandvm]

Finally, something I know a little about.

The big enemy in vet clinics is pet hair. Mount the computer as high as possible to keep hair and dust to a minimum. If your father does not want to change software then the "physical to virtual" idea sounds like an idea you may want to look into.

I will let others comment on actually hardware but one thing you should talk to your dad about is good backups. Very few vet clinics back up frequently enough and even fewer know how (or when) to restore a back up. Set up a clear back up plan and teach the staff how to restore a back up; they should also be instructed on reasons to restore a back and be sure they understand what will be lost (everything since their last back up). I would recommend backing up to thumb drives and have a different drive for each day of the week. Also back up to CD or DVD once a week/month and archive.

I know of one vet clinic w/ a laptop plugged into a UPS w/ a low power printer kept on a cart but not connected to any other computers. When they had a snowstorm collapse the roof in the room w/ their sever one winter they rolled out the cart restored the back up and kept on seeing patients. Some veterinary software is written w/ a version for vets to use in the truck on farm call that then integrates everything back to the main computer when they get back to he office. If your dad is updating software as well, AAHA Trends magazine will have a write up on all the software this summer (june?)

Re:Buy any current workstation and...

[BikeHelmet]

I picked up a fanless VIA Eden board for exactly this reason - running legacy DOS applications for the next X years. I chose fanless so that it won't overheat if the two quiet fans I stuck in there fail.

(With a case)

It's running Ubuntu Linux with VirtualBox installed and Win98SE inside that.

One piece of advice for you - if you have a parallel port licensing dongle for your DOS software, pick virtualization software with "Parallel port passthrough", like VMWare. VirtualBox is free, but last I checked it doesn't support direct access to parallel port devices.

I would think the most likely parts to fail are the HDD, memory, and fans. Pick up at least one backup stick of RAM. I would buy two HDDs, and either RAID them, or have a cron job backup everything at the end of the day.

Go for Hitachi or Western Digital drives. Hitachi ones are supposed to be very quiet. Don't bother with SSDs or flash - they cost too much for what they offer, and may not warn you when data starts getting corrupted. I prefer backup scripts to RAID so that stupid on-board controllers reading SMART data can warn you in big red text upon POSTing. With Hardware+Software RAID, you often don't get that benefit until one drive completely fails.

If these computers are going to be internet connected, make sure you stick a good firewall in front of them. My favourite is a WRT54GL (or similar router) running Tomato or another open firmware varient.

I've only ever had to virtualize single workstations, so I'm not sure how well VirtualBox will work for getting two virtual OS's talking to each other - but I firmly believe this is the safest and most manageable way to go. They don't make a lot of Win95 compatible hardware anymore(if any), and in 15 years, DDR2 memory may not even be available anymore. Virtualized, you can upgrade the hardware if you need to.

Oh yeah - replace your PSUs before they fail, every half-decade or so. It'd suck if the PSU took a stick of RAM with it when it goes. ;) One $30 PSU every 5 years is probably cheaper than one day of downtime when it fails.

Re:Buy any current workstation and...

[Anonymous Coward]

>I have a friend who deals with Vet who has an old Xenix system - they buy parts of ebay in bulk;)

How big a part of ebay?

Or maybe you mean "parts off ebay". Or more correctly "parts from ebay".

Re:Buy any current workstation and...

[Anonymous Coward]

To the submitter: This actually sounds like a fun project. Good luck.

Thanks (i am the submitter), i amd looking forward to some parts as well, even if i do end up buying pre-built systems i will probably do some customizing (dust filters etc..) which i am pretty stoked about, as i like tinkering with hardware.

The only damper on my enthusiasm is that i know it is going to be hard to set up a system for 15 years.

As for virtualisation, that is quite a good suggestion, it would have almost the same hardware -decoupling effect as a webapp rewrite, but without the rewrite..

The only thing that has got me slightly worried is that the DOS software does use some printing functions, and probably IPX networking, so i probably should consider moving the virtual system to XP anyway..

Re:Buy any current workstation and...

[Just Some Guy]

Remember it could be worse: I have a friend who deals with Vet who has an old Xenix system - they buy parts of ebay in bulk;)

Didja know that FreeBSD can run iBCS2 binaries, possibly with the help of library files copied off the current server?

Re:Buy any current workstation and...

[danheretic]

I had the same experience, except that xenix box was running a custom app that the client didn't have the install media/files for. Fortunately it turned out the MB went bad, so I slapped the hard drive into a doorstop 486 I had lying around and it worked fine. Just started cloning the 250MB hard drive as a backup mechanism.

Re:Buy any current workstation and...

[mr_mischief]

If the software that matters is DOS, then FreeDOS under a VM on Linux or XP (or for that matter Vista, Solaris, or OS X) may be an option.

forget it

[Arthurio]

Tell him that replacing the system every 5 years will be cheaper than getting one that will last 15 years. There, problem solved.

Re:forget it

[vajorie]

lol, it's her father, not her client.

Re:forget it

[friendofthenite]

There, problem solved.

That doesn't solve the problem, it ignores his principal request. The guy looking to buy a computer has made it clear that he doesn't want to have to replace his computer every few years; nowhere does the description say that he's looking to minimise cost. Your response is typical of IT (and other) professionals who presume to know users want, rather than listening to what they actually want.

Re:forget it

[rennerik]

Sometimes (most of the time) users don't know what they want, especially when it comes to IT. Many times they ask for the wrong things, and make the wrong decisions. It's our job, as IT consultants, to let them know what the best solution to the problem would be.

For example, asking, "Why do you want your setup to last fifteen years?" may yield answers like, "I don't want to deal with the costs involved with constantly upgrading" or "I'm familiar with my current system, and I am willing to change, but I don't want to have to refamiliarize myself with it every five years" or maybe even "I don't want to have to pay for someone to upgrade our systems every five years."

All of those answers are perfectly reasonable, but all of them are misinformed. It's our job to let them know that, yes, it may sound expensive at first to upgrade every five years, but putting together a bullet-proof system to last fifteen years is much more expensive. We can also explain how to remain compatible (say, via virtualization, as stated in another post) so they don't have to relearn everything every five years. In fact, the experience remains consistent well into the future. And finally, we can say that, again, the costs incurred with upgrading hardware every five years is much less than designing a system that needs to last at least fifteen years... not to mention, the system still has to be maintained, rigorously, so those costs don't just go away simply because the system has been designed to last a longer period of time.

If, after all of that, they're still set in their ways and aren't willing to take your advice, then I suppose you just have to do what they want... but it would be disingenuous for IT professionals to just do what the user asks on spec, because, as said before, most of the time they don't know or understand what they want. There's a reason why we're the professionals and they're not (i.e., I'm not going to tell an architect that I want a house without a foundation and expect him not to tell me I'm an idiot). Why do people insist on doing that when it comes to IT?

Re:forget it

[Anonymous Coward]

Actually,

I have been in IT 25 years now, and used to hear this kind of request all of the time.

You can show him the costs of replacing every five years, versus the cost of building an indesctrutible system.

The reality is hardware quality is not what it was.

The flip side is that 99.9 percent of all doctors usually go with the cheapest equipment around.

The sane solution would be to tell him that 1000 bucks every five years for newer equipment is the saner more cost effective equipment.

Re:forget it

[Hognoxious]

What is this "her" of which you speak?

Re:forget it

[Brad Mace]

Your objection presumes the user *knows* what he wants. They are frequently oblivious to simple realities like the one pointed out by GP. Not wanting to replacing something every few years often stems from the assumption that it will save them money. Your objection is typical of know-it-all users. (Look, I can generalize too)

Re:forget it

[OmanLegend]

I disagree. He's made a request, but what you really need to understand is WHY he has asked for the requirement of 10-15 years. Is it because he doesn't want to learn how to use a new system? (A legitimate concern), or is price his primary concern? Considering the cost of hardware now that might be something that you can educate him about. Sell a solution to your customer's needs, by asking questions, listening and digging into the root need,and in the end they'll probably have a better solution that they realized they could have.

Re:forget it

[brandondes]

You have to presume the user/customer is a moron, unless specifications are given, because most of the time they are.

Re:forget it

[friendofthenite]

You're also focussing on cost as the key issue. There are people around who just want to buy a computer that will sit in the corner and do its job for 10-15 years, just as you might expect from your TV or whatever. If that's absolutely impossible then he'll just have to accept it--but it's not an unreasonable thing to ask for, particularly is it seems to have worked out okay with his existing setup.

Re:forget it

[xenocide2]

>Your response is typical of IT (and other) professionals who presume to know users want, rather than listening to what they actually want.

No, the typical response is to ask why something or other is required, when IT believes the requirement is unreasonable or impossible or undesirable. The typical response to IT asking why people want something is an assault on IT, declaring IT workers never give what people want, even when they've just been asked a deep question about their goals in a very reasonable attempt to understand the people they serve.

Unfortunately, Ask Slashdot rarely provides the opportunity for this back-and-forth. So instead you have to assume some motivation or not answer the question at all.

Re:forget it

[RichardJenkins]

You're right; the first thing to do it understand why he wants a computer system to last 15 years. I suspect that when most business owners understand the cost savings to be had by building it differently they'll choose to go that route.

Some won't. That's their prerogative, but you'd be negligent in your responsibilities as an IT professional if you didn't explore other options with them.

Re:forget it

[Anonymous Coward]

Your response is typical of IT (and other) professionals who presume to know users want, rather than listening to what they actually want.

I couldn't agree with you more. I've been in IT for the last 16 years, working my way up from tech support (still working with dumb-terminals and thick-wire ethernet -- vampire taps anybody?) to network admin, and now a sr. programmer. All too often my colleagues forget that we are here to facilitate the needs of those we serve. Granted, we need to work with them to help them make good, well-informed decisions, but ours is not to dictate to them what it is they want.

Re:forget it

[RichardJenkins]

Two minutes ago I was thinking the exact same thing. Then I realised if a small business owner tells you that 15 years ago he set up a system at minimal cost and is only now looking at changing because the hardware is noticeably ageing, you'd have a hard time explaining to him that it's going to be *harder* to set up a system with similar longevity nowadays.

I wouldn't know how to even approach the subject without sounding like I'm just trying to extract more cash out of him.

Re:forget it

[Anonymous Coward]

There, problem solved.

this is slashdot... why solve the boring problem when we could be solving interesting ones?

Re:forget it

[Anonymous Coward]

My father is a veterinarian

He is cheap many doctors are.

I wrote software for doctors for a few years. They will drop 200k on a bad ass car. But the computer for the front room nurse? "Can we get it for 200 dollars less somehow?". "I found a B/W monitor for 20 dollars less" and so on...

They are running a business. If they can get ROI of years then they are doing pretty good.

I would tell him to buy *THE* cheapest computer they can (350-400 at this point in time). It will crush anything they are currently using. Hell it probably will ROI itself in the power bill alone in the first year. Then in 3-5 years when it goes tits up (and it will because it is a cheap computer) just migrate it.

You are advocating the guy spend thousands of dollars on something that he really doesnt need to do. He probably dropped 5k on those rigs back in 95. He probably can get 15 YEARS worth of computers if he replaces every 3-5. At the same 1995 dollars cost.

Your response is typical of IT (and other) professionals who presume to know users want, rather than listening to what they actually want
Your response is one who doesnt bother to dig into what they are really doing. Then showing what it would cost. *THAT* a business man will listen to. "I can do the same thing for 2000 less but there is a bit of a downside and some risk" will get his attention. Also his business is probably hurting in that floppy drives are long gone. The CD (if it has one) probably is on its last legs.

If he is DEAD set on lasting 15 years then I would say get an asus mini notebook. Plug in some cheap lcds and a 15 dollar keyboard. Probably total cost 500 bucks. Would probably last 15+ years.

15-20 years ago you thought about computers like that. It is not surprising he is still thinking like that. My dad is the same way. Until I laid out the math for him. He can buy a 2500+ computer to surf the web and hold a simple db and print some pages once and awhile. OR buy a cheapo one and upgrade (and keep your computer from becoming legacy junk no one wants to touch) ever few years and keep up and get the exact same effect.

There are other benefits to thinking this way. For example backup solutions. He is probably (if we are lucky) using some sort of tape drive. Older types of tapes are long gone and getting more expensive. But keeping up keeps your costs down.

Computers (least back in 95) were considered a 1 time expense. They are not. They are an ongoing expense. Treat them as such.

When it comes down to it newer computers are cheap junk. Unless you are really willing to pay for the support costs and buy 8k rigs, with long costly service contracts. You have pretty much no guarantee of lifetime. That he got 15 years out of those computers is commendable. But I would not plan my business that way. I would plan as if they are going tits up at any moment. If you treat them that way. Low cost is a better solution.

That he didnt keep up now he has other costs. Such as now he needs to be using VM type solutions or get some sort of conversion done to his data (and hopefully the new software does what his old does).

I would be a poor IT professional if I didnt show him the hidden costs that he has been accruing. There are tradeoffs. It is my job as a IT pro to SHOW management what they are.

Re:forget it

[haruchai]

Bet you still drive a Model T

Re:forget it

[crispytwo]

Agreed.

Anyone that has dealt with users and customers that "know a little" has to help them learn a bit more about reality of hardware. He was lucky that he got 15 years out of it, but that doesn't mean anything, than a pitiful anecdote.

I tell everyone (from experience) that they should plan to replace their business system every 2 to 3 years, and home systems every 5 at most. If they don't plan for that, then they will likely experience a catastrophic failures. Very bad for business, and sad for home users.

This person may discover the hard way that tomorrow, the 15 year old reliable machine will no longer turn on... and can not be rescued. A week later they might have some of the replacement stuff working again, but the pain will be much higher cost than being reasonable in the requests and get a "system" that works long term.

We know 4 things about the person making the request:
1) he has unrealistic expectations for reliability
2) he asking for something that is difficult, if not impossible to guarantee
3) as a dentist, his needs do not match the request (i.e. he is not going on a 15 year isolated excursion to a moon around Jupiter)

We know 2 things about the previous poster
1) prefers to defend the indefensible
2) thinks customers make reasonable requests

I could infer a bunch more.

Re:forget it

[644bd346996]

This guy doesn't want an enterprise IT system. His specs are well below the threshold for an embedded system, and in that market it is perfectly reasonable to ask for a turnkey system that will last for more than a decade. It's quite clear that he'll never be interested in any upgrade, just repairs. Since this system doesn't even connect to the internet, the only reason it would need software updates is if older hardware gets replaced with newer, incompatible hardware.

When you ask "why do you want a system to last 15 years?", the response will probably be "Why not?" If your only answer is that crappy hardware is cheaper, you probably will be fired. This guy is not looking to keep any tech support staff on retainer. He just wants an appliance that just works.

Re:forget it

[crispin_bollocks]

Absolutely true. My beloved 16 year old Micron 486 died recently. Hoping the 1GB Quantum SCSI drive is still okay. This machine cost $4100 - without a monitor, I bought a 21 inch NEC separately for $2500. It was a serious workhorse for 5 years (the monitor for 10), when it was replaced by a machine that cost 1/3 the price. For 500 bucks you should be able to build a kick-ass system that will last him 5 years. If he budgets a maintenance reserve of $10/week, he'll have another screamer in five years.

Re:forget it

[adolf]

I wouldn't know how to even approach the subject without sounding like I'm just trying to extract more cash out of him.

You're still thinking like a frugal geek, not a doctor who just wants stuff to work with minimal hassle.

Of course getting this job done is going to extract cash from him. I'd guess that he's prepared for that; chances are, those systems he bought and paid someone to set up 15 years ago weren't exactly inexpensive, either.

So: Forget money. Please. And just ask: Can it be done?

I think it can.

Myself, I'd use all new parts. A low-power motherboard from VIA or Intel. A good power supply, with high-quality, slow-moving fans, and ball bearings. And I'd try, hard, to cool the CPU passively.

For storage, I'd use a SSD, or a slow-spinning 3.5" drive. I would make sure that intake air passes over (or under) the drive, to keep it at a reasonable temperature. The box would have plenty of RAM (more for an SSD to minimize swap).

And, gosh, I guess that's it. If the machines aren't connected to Teh Intarweb, I'd put the same software (maybe even including Windows 95) back on, and call it a decade.

Re:forget it

[lucas teh geek]

hell, considering the system being replaced and has been good enough until now, you could go dumpster diving every 5 years and upgrade his system for free, forever (or at least until people start disposing of computers in a more environmentally responsible way)

about 10 years ago i got a 1200mhz athlon, it was pretty sweet but within about 3 years it was no good for what I wanted to do with it anymore (was an avid gamer). ram standards changed meaning there wasnt much worth keeping (cpu slow, old style ram, mb didnt support newer processor) so it got turfed. thats a computer with over 10x the power of their current vet system thrown out as scrap. i would of happily given it to anyone who could make use of it. it's not a particularly uncommon scenario either

I would guess the vet is looking at it from an "if it works, dont mess with it" point of view, but from a geek point of view it screams "if anything goes wrong with that pc, you're fucked". good luck finding compatible replacement parts for even a 10 year old pc. remaining current is not just about chasing teh megahurtz. it's about being able to walk into a shop and buy a replacement for whatever component just let out the smoke

Re:forget it

[rts008]

Disclaimer:
I am not in IT, but have worked{and have my degree) in Veterinary Medicine.

I agree with you, but would also like to mention something that I have not seen addressed yet.

While costs play a significant and important argument, future needs also have to be addressed.

Increasingly, computers are being utilized in medicine and medical equipment. Especially controlling medical equipment.

How long is his father going to practice? Long enough that he will 'just have to have' that new fangled lazer surgical bot, or not?
It matters if you plan on no hardware upgrades for 15 years what you implement today.

It's a tough call. The typical small veterinary clinic runs on a slim profit margin these days.

You could save costs now, and find yourself referring your future patients to a better equipped clinic in 5-10 years.

   

Re:forget it

[try_anything]

Yeah, people are quick to push back when the customer asks for something impossible or boring, but not when the customer requests something that is so FUN to geek out about. The fans will go first! Virtualization is the answer! SSDs, so much to argue about there! All so true, all so premature. It's true that the question says, "How do I provide a workstation that lasts 15 years?" but from a customer's point of view, a "workstation" is a working system they can use. You don't use hardware without software, so you can't talk about a workstation lasting 15 years until you know what kind of software you'll be running in 15 years. His father is NOT just asking for hardware that will last that long, or if he is, his question is based on incorrect assumptions that his son needs to discuss with him.

Re:forget it

[Kaboom13]

I wouldn't, because his original solution probably cost a small fortune. Compare PC prices from 1994 to pc prices today. You can probably buy 3 computers today for what each computer cost in 94. The downside of that is they may not last as long. That said, he's damn lucky if all his pc's have kept going for 15 years. In fact, the only reason he hasn't upgraded before is probably because of luck.

That said, while hardware designed to last 15 years is probably extremely cost-prohibitive, you can design the system to make replacing hardware very easy. Who cares about hardware failures if you can drop in a spare in minutes?

Anyways, if you want a system to last a long time with little management, there are some easy steps to take.

1. Use mature technology.
2. Use passive cooling.
3. Provide automated recovery. There will be failures in any system, make it easy to recover from.
4. Document and schedule regular maintenance, with reminders. For example, once a year blow the dust out of the pc's. Clear old entries out of the database. Run a hard drive/memory diagnostics to spot failures before it becomes a major issue.

Re:forget it

[drinkypoo]

If you want someone to hate you for ten years starting five years from now, sell them a computer from today. In five years it will break (or less!) and they'll spend the rest of the time that computer should have been working thinking you're a fucking moron. Nobody builds anything to last any more, and good riddance; the only problem is that we don't build things to be easy enough to recycle.

Re:forget it

[Glonoinha]

As a rule of thumb, I've found two things to be true when dealing with requests of this kind -

1. The doctor probably said '15 years' off the top of his head, as the existing system is his only frame of reference. It was probably said in passing and casually, yet it's been escalated to the most important thing on the list - when in all honesty the most important thing is 'my practice stays in business so I can continue to make money.'

2. It's a good idea to see how much he can spend and adjust the system to that amount, rather than continually down-wrench the system trying to just get the cheapest system humanly possible. Ask the doctor - what did you pay for this system in 1993? Five thousand dollars? So that's my budget, right? If he says no, get a real budget number and make it work, explain the trade-offs for that particular price point.

Re:forget it

[RichardJenkins]

You're right, of course. Not sure where I got that "minimal cost" line from: it's a reasonable assumption that using hardware expected to fail before 15 years could offer savings not possible 15 years ago.

That alone would be reason enough to discuss alternatives with the client.

Re:forget it

[Anonymous Coward]

I wouldn't know how to even approach the subject without sounding like I'm just trying to extract more cash out of him.

You say that like he didn't spend $10K on the old system.
Buy any Dell, get the best warranty program they have then buy a new one when the warranty expires.
Over the period of 15 years you may have to upgrade 3 times. But you'd still be well below the purchase price of the old system.
Hell, 10 years from now who knows how good a low-power solid-state machine might be?

The only hard part would be having to convince him that he was VERY lucky to survive 15 years on the same hardware.
If all else fails, go with the car analogy "Dad, are you still driving the same car you drove 15 years ago? How many parts have you replaced? How much money have you spent keeping it roadworthy?"

To answer the OP's questions:

Do I buy, or build myself?

Buy (see above).

How many spare parts should I keep in reserve?

No spare parts. Computer parts aren't known for their shelf-life. Just call the manufacture.

What will fail first, and how many years down the line will that happen?

It depends on the usage/environment.
Do you have conditioned power? A nice cool place for the computer? Lots of dust (or hair being a Vet)?
Hard drives and fans go first (3-5 years), then power supplies and capacitors (4-8 years), all should be covered by a decent warranty.

The other question, do you want to be the support guy for the system for the next 15 years? Or would it be nice to tell dad to "just call dell"?
The only other advice would be "Backups!"; environments like this typically don't have disaster recovery plans.

Re:forget it

[DerekLyons]

That doesn't solve the problem, it ignores his principal request. The guy looking to buy a computer has made it clear that he doesn't want to have to replace his computer every few years

My nine year old niece wants a pink pony with wings, she isn't getting one any more than this guy is getting a computer that's going to last a guaranteed 10-15 years. Both the pony and the computer are imaginary animals.
 
 

Your response is typical of IT (and other) professionals who presume to know users want, rather than listening to what they actually want.

No, his response in one of someone used to dealing with reality rather than playing marketing drone and promising the customer something that can't possibly be delivered. The customer isn't always right.

Re:forget it

[DerekLyons]

Then I realised if a small business owner tells you that 15 years ago he set up a system at minimal cost and is only now looking at changing because the hardware is noticeably ageing, you'd have a hard time explaining to him that it's going to be *harder* to set up a system with similar longevity nowadays.

Nowadays is no different than back then - especially because I doubt the guy spent any money on getting a high end ultra-reliable system. He's probably got a straight vanilla commercial box.
 
Now, because he got lucky and got a system way the hell out on the right end of the bell curve, he mistakenly believes that such things can simply be ordered off the shelf.

Re:forget it

[MacTO]

While it is fair to consider yourself (as a presumed IT professional) an expert in the hardware and software that he runs, you have to keep in mind that you are not an expert in the operations of his business. There are a variety of reasons why this person may want a long lived system. Hardware, software, and retraining costs may be among those reasons. Then again, he may have some other factors in mind. I can't presume what those implications are because I am not the one operating his business. On the whole, I do agree with the assessment that it is not reasonable to expect modern hardware to operate for 10 to 15 years. And I do agree that virtualization is probably the best solution given the customers demands, since that will allow him to run the same software that he has run for the past 15 years into the indefinite future. After all, people usually care more about their interaction with software than they care about their interactions with the hardware. So it may be worthwhile asking the customer if that is an option in his mind.

Re:forget it

[gad_zuki!]

If the request is irrational then ignoring it is the right thing to do.

There was a website a few years back of irrational project requirements for web designers. Stuff like "DVD quality full screen video with 0 wait time or buffering."

End users dont understand that 15 years for one computer is a rarity, 10 too, and 5 being more practical. They dont know what RAID is nor to they understand the importance of backups. Teaching them these things is important. Teaching them the limitations of hard drives and power supplies is the ethical thing to do.

All specialists have to deal with irrational and ignorant customers. You serve them best by educating them. You steal from them when you sell them snake and oil and make bullshit promises like "this will last 15 years."

Re:forget it

[Cyner]

Thank you! My first thought was "hardware that lasts 15 years is more than 3x more expensive than hardware that lasts 5 years.

This is a business decision, pay attention to the TCO (Total Cost of Ownership).

Re:forget it

[Anonymous Coward]

You tell him the system has lasted 15 years by chance. I would suggest that he could purchase another system, and hope it lasts 15 years, but with old hardware - not only does it become more likely that it will fail, but it also becomes harder to get reliable replacement parts and as such, the downtime will be much longer.
We have a customer on an 8 year old server with 50 users. We have been telling them for the last 4 years to replace the hardware, and they don't listen. I then asked them how much it would cost them if the server died, and replacements couldn't be sourced for two days. They are now provisioning a new server.
So the question you should put to your dad is "How long can you do without a computer system?"
That will give you a basis of how often you backup, and backup your backups offsite, how often you replace your machines, and how you plan for provisioning new hardware in case of failures.

Re:forget it

[Anonymous Coward]

Simple:

"You got lucky. Three parts of the computer fail regularly: Power supplies, Hard Drives, and Fans. Ensuring that each of these has a probability to last 15 years is significantly more expensive than getting a setup that SHOULD last AT LEAST 5 years. Oh, and make back-ups, in case it doesn't."

Anything you make SHOULD have the ABILITY to last 15 years. Whether it's PROBABLE or not is another matter. He beat the odds before with an inexpensive-ish machine. Odds of him beating the odds again are, well, against the odds. So, either pony up or play the odds again.

Re:forget it

[adolf]

Bah.

I've got computers at home and at work which have lived for more than 5 years without any real issues.

I look forward to the day when buying a computer for a specific application is like buying a circuit breaker panel: Once it's properly installed, it will do what it's supposed to do pretty much forever (in human terms, at least).

Re:forget it

[couchslug]

Buy multiple machines and spare parts. Box them up and put in service as needed. The logical goal is a fifteen-year supply of the same machine and OS.

Re:forget it

[hedwards]

I understand the problem, but which would you rather be dishonest or greedy? Ultimately, 15 years is a long time even for a well cared for machine to last. Realistically you're not likely to have somebody paying you to do it for them that could do that kind of work.

Best case you get really lucky, you put the machine together and the hardware lasts for time. Likelihood of it happening without a lot of maintenance, negligible. Likelihood of somebody having to be paid to keep the ancient pile of crud going from years 10-15 almost certain.

Yes, you do run a very high risk of looking greedy. But at the same time if you don't he'll likely pay through the nose in the short term, pay too much for maintenance later, and probably pay way too much when the system abruptly dies and can't be resuscitated.

So yes, it's a hard sell and it makes you look kind of bad, tactifully pointing out the extra costs involved with the demand over just replacing the hardware is probably the best bet. With the upside being that you don't have to spend as much time setting it up or charge as much as you would otherwise.

Re:forget it

[mgblst]

Look, anyone with any sense knows they don't make stuff like they used to.

Re:forget it

[nine-times]

Then I realised if a small business owner tells you that 15 years ago he set up a system at minimal cost and is only now looking at changing because the hardware is noticeably ageing...I wouldn't know how to even approach the subject...

Maybe by pointing out to that business owner that he probably doesn't see lots of other 15 year-old computers around. It's really not just because of disposable consumer culture.

Not knowing the actual failure rates is over 15 years, I'm sure it's really high among both computers manufactured 15 years ago and computers manufactured today. So if you're the guy whose computer is still running after 15 years, then it's not really so much because you made a good purchasing decision. You got lucky.

So taking that information to an IT pro and asking, "How do I buy another computer that will last 15 years?" is a little like winning the lottery and then going to an accountant and asking, "So how do I buy another winning ticket?"

Re:forget it

[Anonymous Coward]

Can I suggest you introduce a significant off-site backup system to this solution? If the facility burns down, there is very little in the way of backup.

One point I might make - When you build new computers, you should consider the licensing you are using - Microsoft will not sell MS-DOS to you, you can't buy a license for it and you can't transfer licenses from MS-DOS 5x onwards. How are you going to get around this? Do it illegally?

Additionally, Why not change the hardware over? What is his business worth if a computer breaks down and he can't recover records? He can write computer purchases off on tax, by depreciation, so the equipment isn't really worth over-engineering for 15 years - Manufacturers accordingly do not expect domestic-grade hardware to be operational in 15 years time - they build to an 6-8 year lifecycle with the expectation that you will have been finished with it within 3-4.

To get 15 years these days, you'd be pushing the proverbial shoit up hill, unless you buy server-grade components. I'm not saying don't try - I'm implying it may not be worth it - the customer may actually be wrong here and you haven't the ability to say no to it due to the relationship.

Re:forget it

[Sj0]

Ew, no fans.

I deal with reliability on a daily basis, and mechanical parts are a pain in the butt. They need huge amounts of maintenance to be able to give any dependability. If you can design without fans, then your system will be more reliable.

It'd take more engineering, but utilising convection currents to move air instead of fans will mean fewer components, which will mean greater reliability.

Re:forget it

[Sj0]

If you got the secretary's PC running the latest whiz-bang medical laser, then I hope your malpractice insurance is paid for.

Re:forget it

[laird]

"I realised if a small business owner tells you that 15 years ago he set up a system at minimal cost and is only now looking at changing because the hardware is noticeably ageing, you'd have a hard time explaining to him that it's going to be *harder* to set up a system with similar longevity nowadays."

A computer system 15 years ago might have cost around $2,000, which is over $2,800 now. You can get a decent PC now for $500. This drop in price is achieved, in part, by reducing the quality of the engineering. For example, consumer PC's now have much worse power supplies, components that can't be repaired, etc., all of which drives down purchase price.

It's probably worth digging into why he hasn't upgraded in 15 years. Does he really hate spending money? Or does he think that because his 1994 computer was hard to learn (Windows 3.1, crappy networking, etc.) that he'd have to go through the same thing again? This may sound silly, but I've known many older folks that stuck with truly horrible software for years because it took them forever to learn that software, and they think that any new software will require as much effort. But if you stick them in front of a modern machine, with modern software, they're fine, because in reality modern software is much easier to use than software from 15 years ago.

I'd second (third?) the suggestion of using virtualization. That way any modern PC can look exactly like his museum-pieces, and (assuming he's got backups) he can swap PC's without any changes.

Re:forget it

[Sj0]

I'm sort of shocked that nobody has stated the obvious:

You want your system to run for 15 years without problems, because it's a massive pain in the ass having to deal with your computer system being down, so having a reliable system will mean you minimize the chance that you'll have to deal with such a pain in the ass.

If I was running a business, and I had a device which run perfectly fine for 15 years while the rest of the world whined about their systems going down, I'd want to continue to not have to worry about my computer systems.

It's cheaper to buy a yugo until it breaks then replace it than to buy a reliable vehicle once. It doesn't mean the former isn't a massive pain in the ass.

Re:forget it

[rts008]

I would hope so too, but just drop the clueless attack. It just shows you know little to nothing about small veterinary clinics.

Not likely to be a widespread problem though, as small clinics rarely hire secretaries....never mind a secretary's computer. Most likely you will find one of(or the only one) the Vet.'s Vet Techs acting as receptionist, secretary, and and Dr.'s assistant.

Oh, and the Vet Tech is probably the one taking care of all of the tools/equipment in the clinic anyhow, including the 'latest whiz-bang medical laser', and will be more familiar with it than the Dr. is.

Better luck next time!

Re:forget it

[Sj0]

I do control engineering, and I know IT well. Neither field would consider a piece of medical control equipment an appropriate device to multi-purpose.

You're begging for trouble. It's not a safe way to cut costs. If I suggested cutting costs with such a multi-purpose system, I'd be putting my career and professional designation on the line, and if the medical profession has any sort of standards, so would any doctor who requested it against my advise.

Notice there have never been any PLC/internet appliance combos? There's a damn good reason.

Re:forget it

[arkane1234]

It ranks right up there with people asking for Windows that doesn't have viruses or applications that crash.
15 years ago = 1994
A low end 486 was common then... Anyone expecting (that's the keyword) that hardware to last 15 years has no idea what anything is about with the hardware and needs to be educated a touch to make that kind of decision. Otherwise, if the education part is laughed at, the person will have to consult with someone who knows and take their recommendation instead.
It's pretty common in this world. We go to doctors because we don't have the time or ambition to learn it. We go to auto shops because we don't have the time or ambition to learn it. Now of course that doesn't apply to everyone, but it's just how society works in general.

Re:forget it

[supernova_hq]

Yes, 5 years ago they made things to last. Now they don't!

And if you are still waiting for parts that last pretty much forever, keep waiting...

Re:forget it

[Anonymous Coward]

"For example, once a year blow the dust out of the pc's." Once a year, at a Vetrinary Hospital? This would be one reason why it fails sooner. Heat.

Re:forget it

[lapagecp]

The guy looking to buy a computer is not IT savy at all. The problem is that the customer has not told the IT Pro what he wants. He told the IT Pro what he thinks will solve his problem. Most people understand the concept I am talking about better when it comes to health care. You don't go to your doctor and say I want Lipitor. If you do, your doctor is not just going prescribe you your drug of choice. You tell your doctor your problem and what you want for an outcome. Then the doctor helps you decide. A client asking for a computer that lasts 15 years is equal wrong. The answer is always going to be, that's impossible, what are your concerns, what is the problem you are trying to solve. Now lets think of a solution that comes as close to that mark as possible. If your client wants a computer that will last 15 years and you give them a computer that could last 15 years you are not doing your job. If you have a computer that will last 15 years then your client can not afford it. I can easily give a client a computer that will last a long long time and be really easy to fix should something go wrong.

Re:forget it

[sp3cialk79]

I'm still running as 8086 IBM PC, with 64k mem, 20MB hard drive. Just had to replace fans, the hard drives I've found at my local flee market or reuse centre. I use it for my accounting software, too lazy to replace it or migrate it. It aint broke so why fix it.

Re:forget it

[ujoronen]

...they do this in IT because they honestly see IT pros as kid geniuses you can pay off in twinkies and a pat on the head.

I see this all the time even though I am considerably older than the average age of my clients.

The other ones consider us part of overhead, since we aren't in sales, even though without us they really couldn't even spell CRM. Oh, and please don't mod this one +funny, because it's really not.

Re:forget it

[Just Some Guy]

Two minutes ago I was thinking the exact same thing. Then I realised if a small business owner tells you that 15 years ago he set up a system at minimal cost and is only now looking at changing because the hardware is noticeably ageing, you'd have a hard time explaining to him that it's going to be *harder* to set up a system with similar longevity nowadays.

"You have a '68 Mustang and it still runs. What would you buy today that will still run 40 years from now?"

Car analogy FTW.

Re:forget it

[DigitalCrackPipe]

Use televisions or other modern electronics as an example. While televisions produced a few decades ago lasted decades, newer components are designed with the explicit expectation that they will only last a few years. You simply can't buy very many products with the same longevity anymore, and expecting to have the exact same experience is setting yourseslf up for disappointment.

Re:forget it

[mr_mischief]

The car analogy might just fail. With proper maintenance and a garage for storage, you can drive the same car for twenty or thirty years. With some luck and some extra maintenance, we have running cars around from a century ago. Sure, most people let their cars undergo some extra wear and trade for a newer model, but not everyone does.

I don't get rid of a car until the maintenance becomes more expensive than the payments on a newer car or someone totals my car out on the road. I almost never buy brand new even then. Let someone who cares about new car smell pay the first year of depreciation.

My gaming rig at home is getting long in the tooth, so I'll probably build something newer in a few months from parts a generation or so old from NewEgg. My work PCs, though, don't have to do much more now than five years ago. This post is from an Athlon 1000, which might soon get upgraded all the way to an Athlon XP 2400+ or a Sempron 2800 and its Socket 754 motherboard pulled from another machine. The main problem with this PC isn't the processor, but the cost of adding more PC-133 RAM rather than replacing the motherboard and using cheaper DDR or DDR2.

Re:forget it

[mrbooze]

I feel like my response would be something like "Well, I've never in my life put together a standalone computer system that lasted fifteen years without maintenance, but you've already done it once, so why are you asking *me* for advice. Sounds like I should be asking *you*."

Re:forget it

[Anonymous Coward]

Your response is typical of IT (and other) professionals who presume to know users want, rather than listening to what they actually want.

Your lips were moving, but all I heard was "blah blah blah install Ubuntu Linux blah".

Re:forget it

[Anonymous Coward]

Use a laptop? What are you smoking, Ricotta crack??

Build yourself

[drachenstern]

Expect the fans to go out first, then the power supply. It wouldn't hurt to build a duplicate of the system, for spares; however since that defeats the purpose of the build a single box strategy, then obviously that won't work so well.

What are you thinking for storage? I would at the least focus on SATA (the 3.0 spec) and use probably software raid, so you're not stuck on a hardware raid failure causing ultimate data loss. This is probably one of the only times in my life I have ever suggested software raid, but since you have low processing requirements, this should not be a problem.

I think that people are going to say hard drives, but those wouldn't be the first ones to go out. And what about virtualization options? That lets you move around to various hardware without causing an upset on the system build. You might look there, and then even if you replace the base system every couple of years there's no upset, except data retention (meaning, make sure you're running RAID 1 on the boxes...)

Re:Build yourself

[Anonymous Coward]

When exactly software RAID is not the best option? I mean, any modern CPU can XOR several gigabytes per second... it's unlikely any RAID card can beat that. Of course you need good bandwidth from the card to CPU. Times have run past hardware RAID anyways...

Re:Build yourself

[timeOday]

use probably software raid, so you're not stuck on a hardware raid failure causing ultimate data loss

Software RAID won't boot if the primary hdd fails, will it?

Re:Build yourself

[Anonymous Coward]

Agreed with the fans. His original setup, a 486 and a pentium, probably didn't have cpu fans. My memory may fault me, but I don't think the 486 even required a heatsink.

Re:Build yourself

[hawkinspeter]

You can setup grub to boot a fallback partition if the primary partition isn't available - no problem at all in software. Software raid gets round the problem of having to find a compatible raid controller if the controller fails - that can be a problem if the machine is 10 years old and you're running some old bus technology.

The only downside of software raid is that the machine stops if a drive fails and it's not hot-pluggable, but a quick power cycle should solve that problem (provided you set up grub properly on both disks).

Re:Build yourself

[timeOday]

You can setup grub to boot a fallback partition if the primary partition isn't available - no problem at all in software.

But how does grub get loaded from the boot sector of drive 0 in the first place if the drive is broken?

Re:Build yourself

[Cyberax]

It will (assuming you are talking about RAID1). You just need to install MBR on both drives.

Re:Build yourself

[vlm]

If SW raid 5, disregard rest of my post, I think...

If the bios says its totally dead, it'll skip to the second. No problem.

Or, if using mirroring, and removable drive caddies, swap the drives or simply pull the primary. That'll make it switch to the "second" drive for certain.

I've been doing this for years, and yes I have successfully pulled both drives (not at the same time of course). I also had a drive failure, and successfully fixed it this way.

If you forget to install grub on the second drive, of course, it doesn't work so well.

Re:Build yourself

[Enigma2175]

But how does grub get loaded from the boot sector of drive 0 in the first place if the drive is broken?

IIRC, if you are using a RAID 1, Grub will write the boot block to both disks.

Re:Build yourself

[hawkinspeter]

You load grub on both boot sectors, so grub will be loaded from the boot sector of drive 1 when drive 0 isn't available.

Re:Build yourself

[hawkinspeter]

When I've set it up, I've had to write grub to the secondary boot sector manually, but I don't normally set the boot partition to be RAID1 as it's tricky to get grub to boot from a RAID - I load the modules in the initrd.

Re:Build yourself

[drachenstern]

Not that I should bother replying to an A/C, but it did only on some models require a fan. However, I've yet to see a 486 that didn't omit the heatsink. They've always been quite necessary. The huge chunk of crumbly stuff on top of older CPUs (ya know, the part that _held_ the silicon) was effective, but after a couple hundred thousand transistors within a sufficiently small space at that voltage, a heatsink is pretty well mandatory.

And most boxes did have fans at the time, but you're correct, likely not cpu fans.

Re:Build yourself

[SPY_jmr1]

The 486 might be passive, but i've not seen a passive pentium before. Perhaps a later slot 1 type celeron or P2, maybe.

In fact, the original 5v pentium (think 60mhz) was one of the hottest running cpu's i've ever seen, before or since.

Thinking back on it, it probably seemed much hotter then it was on an absolute scale. The heatsyncs and fans used on those old things were tiny compared to ones on a modern cpu. But still, they ran quite hot in any case (the pentiums).

Re:Build yourself

[drachenstern]

Um, I think I remember seeing some of the first generation be fanless, but the cases still had fans. After that first set of them though, fans became pretty well mandatory. But my comment about passivity was in regards to 486's, not Pentium class, so take that for what it's worth.

I remember all the Cyrix chips came with HSFs that I ever saw, but those were usually upgrade chips, so they naturally were beefier and therefore were of a faster speed to begin with.

Re:Build yourself

[SPY_jmr1]

Oh case fans. Yes, certainly.

The AT class machines, more specifically, power supplies (think the kind with the 'big red switch', which would still be awesome to have, but I digress) usually had a fairly hefty fan on them.
That combined with case design probably provided for a decent amount of cooling on its own.

short answer - you don't

[petes_PoV]

You cannot guarantee that lifetime, so the best response is to design a flexible solution. One that *could* last that period of tim if there are no hardware problems.

However, you should consider how to upgrade each part in isolation - or with small numbers of associated changes. That means using popular, but not bleeding-edge components. One's that (like with vintage cars) have a good number of enthusiasts using them. That means that spares will be available and the know-how to diagnose and fix problems will be available too.

The final fallback would be to buy two systems. Keep one in "deep freeze" until you need to cannabilse it for spares. However, don't expect the electrolytic capacitors to last that long.

Re:short answer - you don't

[MouseR]

Or just get quality components to begin with.

At the office, I'm still running a 350mghz PowerMac G4 computer (the bugger is 10 years old) as a server.

All original components. None failed. System still has it's original bleeding-edge 320megs of Ram, runs Mac OS X Tiger.

It was given a new 40gig baracuda drive that's been sitting on shelves for years. had never been used.

We use this machine as a slowest-denominator software test platform for a product in development and as a distributed networked compiler farm node and backup server for another more important machine (it backs up the backup machine's main OS, not it's files).

MS can argue all it wants about Apple making "aesthetic" machines, they actually use good components. Current XServe hardware being another case in point.

Re:short answer - you don't

[afabbro]

At the office, I'm still running a 350mghz PowerMac G4 computer (the bugger is 10 years old) as a server.

Hmmm, if that's mghz = MegaGigaHertz, then I'm quite awed. But if it's MicroGigaHertz, then I feel bad for you.

Re:short answer - you don't

[SatanicPuppy]

If you have extremely modest needs, you can get by with extremely modest hardware. For Tiger, that system is butting right up against the ram requirements, and forget getting new parts...With the intel switch all the original mac parts are collectors items, and extremely expensive...I had to replace a wireless card on a newer machine recently and it cost nearly 400 dollars.

And frankly, I'd never recommend using anything but a new drive on a system you care about. Old drives don't store well.

Re:short answer - you don't

[nine-times]

Well just to take a step back, the big question is, what is the reason for the requirement that the setup must last 15 years? Is it cost? Is it an aversion to change? Is it some sort of ideology opposed to the culture of disposable goods?

Any system can last 15 years, supposing it continues to meet the user's needs and assuming nothing breaks. There are limits to how well you can predict those things, though. Something new might be available in 5 years that would be extremely helpful for the user, in which case it becomes worthwhile to upgrade. Also, even with the best quality products, some percentage of units will break within 5 years. The likelihood of something breaking in 15 years is pretty good.

It's also worth asking why the current system can't be repaired/upgraded. If his 66MHz DOS system is working, you might be able to get a bigger hard drive that's compatible, install Windows 95/98 on it, and keep going. I know it's not sexy, but if the point is to stay cheap or avoid change, and the current system is doing its job, then why not?

But ok, let's say you want to get a new system. What are the requirements? Does it need to run the same DOS application? Have you tested that application in newer versions of Windows? What about WINE? If it doesn't need to be the same DOS application, what's the new application going to be like? There are lots of questions.

As far as hardware, I would say the best thing is to buy a midrange business-class workstation (e.g. Dell Optiplex) that seems like it will meet your needs. Maybe make sure you have a couple extra expansion slots or something, in case you need them. There's a non-zero chance that something will break, but you can't stop that. By getting a midrange business desktop from a major vendor, there's a pretty good chance that you'll be able to find support and replacement parts if you need them, which is really the best you can do.

As far as software, use open source if it's an option. Not so much for ideological issues, but over the course of 15 years, you'll be more likely to find an upgrade path with support for legacy applications and hardware if you're running an open source OS. And that upgrade path will probably be free, if cost is the issue. You really don't want to keep someone running a 15 year-old unsupported operating system on their desktop if you can avoid it-- not in the Internet age. It's much better if you can at least get security updates, even if you're not interested in new features.

Re:short answer - you don't

[NormalVisual]

Even before the Intel switch, some of the PowerMac parts were ridiculously priced. I've mentioned it before, but I have a dual-proc G4-450 that sits uselessly in a corner because the proprietary power supply died a couple of years ago, and for what that power supply would cost to replace, I could build a PC that would run a hacked OS X image in a VM at least as well as the old G4 did.

Re:short answer - you don't

[Anonymous Coward]

Actually..old drives are most reliable. i have a 1 gig quantum fireball drive which lives for more than 10 years now in top shape with no bad sectors.

but seriously, the best way is to keep all business records in portale formats like documents, databases, etc.. that way it will not rely much on software and hardware.
you can upgrade your hardware every few years and convert your data into existing formats as you go.

Re:short answer - you don't

[Anonymous Coward]

If you really spent $400 on a wireless card that everyone else sells for not-$400, then you should probably do a better job of sourcing your parts in the future. (Even the original, 802.11b, hasn't-been-made-in-6-years AirPort isn't more than $200, unless you're either an idiot or not in the US. Even the new N cards, which you can't use in a PowerPC Mac, aren't $400.) There's a ton of parts out there, and a ton of upgradability - moreso than for PCs of similar vintage, I think; you can shove a 1.4GHz G4 in that G4/350 if you want to, and up to 2GB RAM, and for a lot of the parts (memory, hard drive and optical drives), they're the same stuff you'd get for a PC of similar vintage.

Re:short answer - you don't

[Anonymous Coward]

With the intel switch all the original mac parts are collectors items, and extremely expensive...I had to replace a wireless card on a newer machine recently and it cost nearly 400 dollars.

Utter B.S. -- Used parts for old Macs are dirt cheap (giveaway items, in fact) and there is no difference between the wireless network cards for intel vs ppc. In fact, I just bought an airport extreme from Apple brand-new for $59.99. The old Apple stuff pretty much works forever; I have an old 68k-powered SE that's never been opened except to replace the battery. Still running its original install of System 6.0.8, too.

Re:short answer - you don't

[Anonymous Coward]

"MS can argue all it wants about Apple making "aesthetic" machines, they actually use good components."

WTF!?!? Quality components die all the time with less than a few hours usage on them. Or is this simply some Mac fanboi's way of trying to smugly say he thinks he has a superior piece of hardware?

Re:short answer - you don't

[cbiltcliffe]

Or just get quality components to begin with.

At the office, I'm still running a 350mghz PowerMac G4 computer (the bugger is 10 years old) as a server.

I fully agree with this. I'm running a Pentium 166 with 48MB RAM and NT 4.0 on my network as an answering machine/fax server. (Yes, I could run it with Linux, but I just haven't taken the time to migrate it.)

I put it together myself from scrap parts that I had kicking around, and it's basically been set up and ignored ever since. Although it's probably overdue to pull the cover and clean the dust out.

My philosophy has always been "If it does the job you need it to, don't mess with it."
That's not to say you shouldn't install security updates and the like, but there's no reason for this machine to move to more powerful hardware, as long as what it's got still works.
If it blows up, I'll upgrade. Until then, it works just fine, thank you.

Re:short answer - you don't

[aschran]

Lowercase "m" means "milli," as in "milli-gigahertz." Which is the same as one megahertz. So actually, he is inadvertently correct. :)

Re:short answer - you don't

[Anonymous Coward]

Oh pashaw. My 75mhz monorail is still running. It even still has the original windows 95A install on it.

All i use it for is playing mp3's off the network.... its been serving this purpose almost non-stop for nearly 10 years... not bad for a machine i picked up used for $20.

Re:short answer - you don't

[Lord Kano]

MS can argue all it wants about Apple making "aesthetic" machines, they actually use good components. Current XServe hardware being another case in point.

I'm a former Mac user. I also worked for years as a repair tech. Until about 5-6 years ago, I STILL had an original Mac Plus. A machine from 1986 worked just fine up until I gave it away in like 2004. I also remember those summer 2000 iMacs, those things were blowing analog/power boards like popcorn. Apple was pretty good about fixing them, but every once in a while Apple makes a piece of crap too.

I still have a couple of 6100s that still work just fine. These are what, 16 or so years old? Those old machines were alot simpler. They produced less heat so the fans didn't need to spin as fast. Slower fans are longer lived fans.

LK

Re:short answer - you don't

[Anonymous Coward]

You sure it's not milligram Hertz? This baby is quite a pill popper.

Re:short answer - you don't

[Anonymous Coward]

And you're lucky. Many MANY *MANY* Mac models do not last that long (virtually every newer G4 and G5 mac model, really). The B&W G3 and early G4 range were very reliable.

          That said, it does throw one point toward older hardware -- if it's already been running >=5 years, you probably don't have an EMachine, Packard Bell, etc. with bad fans, bad caps,faulty/undersized power supply, etc. or it would have died already.

Re:short answer - you don't

[PMBjornerud]

Lowercase "m" means "milli," as in "milli-gigahertz." Which is the same as one megahertz. So actually, he is inadvertently correct. :)

But you must be consistent, which means that "g" isn't giga. It's 9.8 m/s^2.

Even factoring in the milli, running at 0,0098m/s^2 for 10 years is damned impressive, the computer should be up to 1% of light speed by now!

...unfortunately, he also wrote "hz" instead of "Hz".

Re:short answer - you don't

[Anonymous Coward]

Lowercase "m" means "milli," as in "milli-gigahertz." Which is the same as one megahertz. So actually, he is inadvertently correct. :)

Well no, not really. "Milli" is not used in that context. That would mean that "50mbps" means 50 millibits per second, according to you...

Re:short answer - you don't

[hasdikarlsam]

It does. It may be hard to /transmit/ fractional bits, but they still /exist/, and such a measurement would be quite reasonable for describing eg. a really slow compression algorithm.

Re:short answer - you don't

[Anonymous Coward]

It's most likely MilliGigaHertz, which translates to MegaHertz.

Re:short answer - you don't

[toddestan]

I wouldn't count on newer Macs lasting that long. The level of quality has gone down, and Apple is running them way too hot to expect them to last anywhere near 15 years. Furthermore, with the exception of the Mac Pro, they are not built with repair in mind and if anything other than the RAM and probably the harddrive goes bad you'll probably end up having to trash them once the warranty is over. Even the Mac Pro would be a bit iffy after a few years as the power supply/motherboard are non-standard parts and once Apple drops support for them the only place to get a replacement would be from a donor machine.

Your best bet would be to go with a low power, well cooled PC system with high quality, standard off-the-shelf parts. Use common parts, stay away from stuff that will be hard to come by in a few years like BTX formfactor stuff or a hardware raid controller. With a bit of luck, they'll easily last 3-5 years. In that time, when everyone else is upgrading to new machines, save some of the better machines like the ones you built from the scrap heap and throw them in a closet so you'll bave parts to keep the main machines going.

Re:short answer - you don't

[infalliable]

What you need to consider is that the requirements on the older technology were less.

The hard drives spin slower, so the parts wear more slowly. The data densities are lower, so entropy flipping bits is less likely.

The power requirements are lower, so less heat and less diffusion and less wear.

All things being equal, larger feature sizes are more fault tolerant assuming there is no critical initial fault.

Build your own system. Keep it solid and simple.

[philibob]

Moving parts usually fail first; get some solid state hard drives. Avoid fans by using components with passive cooling; most importantly get an integrated video card and a passively cooled power supply. Running DOS software? Use freedos. No need to bother with full-blown Windows. Keep to name brand components and you should be fine. As long as you buy standard components, they should be easy enough to replace 15 years from now, don't go hog-wild stocking up on replacement parts. Good luck!

System Longevity

[jimbo1708]

Wow, that type of longevity is very ambitious. If its just hardware he wants upgraded, then go for it. If its software, then leave it to the pros because there are plenty of canned office management suites out there. Yes they are expensive, but you know that if you install, then the next 10-15 years, you will be the tech support!

virtualization = future-proof.

[spazdor]

If the goal is to get legacy DOS software running on new hardware and being robust, then the most rock-solid option (and maybe the cheapest) will be to put it into a VM such as qemu or VMware. This will allow you to transplant it to new hardware, make/restore backups, far more seamlessly.

As for the hardware itself, have you considered a Soekris box or similar?

Re:virtualization = future-proof.

[corychristison]

VirtualBox is cross platform and open source (used to be Sun xVM Product).
It's really nice. The (Qt-based) UI could use a bit of work, but I've found it to be much, much faster and more stable than VMware Workstation(v4-v6). My use was WinXP Pro on Gentoo mostly for Photoshop CS2... but it's now running flawlessly under WINE so it doesn't get used much anymore. Still use AnyDVD HD every so often but it's becoming less and less frequent.

Industrial PCs?

[Rorschach1]

How about industrial hardware? You'll probably pay at least twice as much as you would for a consumer desktop, but PCs made for industrial control applications tend to be a lot more rugged and build to serve for many years in harsh conditions. Sounds like you don't need a lot of processing power, so you could probably get by with a fanless system and eliminate a major failure (and noise) source.

I haven't bought anything from these guys, so I don't personally know anything about their quality, but SuperLogics has a barebones fanless Atom-based system for $315. Something like that might be a good start.

Re:Industrial PCs?

[maxxard]

I'd go with this idea for the hardware. Industrial Atom based PC with no fan and running of a single 12v power supply - it will be easy to get 12v for some years yet. Something like the (soon to be available) Intel D945GSEJT would do the trick. That shouldn't be very expensive either, so buy spares. On the software front, if you can get what you want to run on a suitable linux distribution, you could burn a bootable image of the system to CD or DVD, and use that from CD, DVD, USB stick or even in a virtual machine if the software you choose isn't too heavy on resources on an Atom based board, plus doing that regularly will sort out backups. Pick a linux distro which is going to be around for a long time (Debian anyone?) and application software which stores things in an open format which you'll be able to read in some years time. Peripherals you're going to have to treat as consumables, as none are built to last 15 years nowadays unless you pay big money, but USB is likely to still be around in one form or another for quite a while. If USB3 ends up backwardly compatible, that should see out several years.

Re:Industrial PCs?

[Anonymous Coward]

In that vein, I can recommend the slower range of VIA mini ITX boards. I've had an 800mhz one running constantly for 6 years. And get this, the fan failed and I didn't replace it for 6 months! Still going.
Mount it well and you have something that is extremely abuse tolerant.

Re:Industrial PCs?

[Anonymous Coward]

but PCs made for industrial control applications tend to be a lot more rugged and build to serve for many years in harsh conditions.

This is a good idea since an industrial PC is probably better protected from dust and the blows of a particulary nervous customer (hopefully not of the owner of the customer).

Re:Industrial PCs?

[Animats]

There are, in fact, industrial PCs designed for long life and support over long periods. It's not at all unusual to expect industrial electronics to last 15 years. Welcome to the world of underclocking. In the industrial world, everything is oversized and derated.

Re:Industrial PCs?

[Anonymous Coward]

Do you think his current setup is an industrialized one? I suspect the "client" here got a basic solution for them and stretched its use for a few years more than expected.
there's nothing wrong with that as long as it works. If they want something more capable and careful planning, good, but this guy would probably not be upset if this upgrade opportuity consisted exactly on doing what they did 15 years ago: get bog-standard hardware and make it run.

specify...

[Anonymous Coward]

if you're planning on replacing the software itself, along with the hardware - don't. i've gotten my veterinary software (DVM Manager DOS) to run on everything from DOS to the Windows 7 beta. If you're just replacing the hardware, then along with a bit of extraneous information, you're pretty much just going to want to go with top of the line business-class pc's. considering that the software you're probably running precedes the hardware by a decade or so, you shouldn't have any issues for a looong time.

Mainframe

[Anonymous Coward]

just buy your dad a mainframe. The knowledge you learn could also help you get a job at IBM in -20 years.

Neurosine

[neurosine]

Go with an entry level enterprise server, like a Dell T300w/ 5 Yr. NBD warranty, throw on Xenserver 5, Spring for 2003 Standard, and possibly 2008. Install both OS's, develop for 2003 with the idea of migrating to 2008 as you can run both OS's live and migrate at your leisure. At the other end place a Wyse terminal(or use the current workstation as an RDP client if it's not too flakey.) With an ADSL connection he can have a consistent environment from work or home, and more terminals can be added as necesary. I know you could have done this for less than 5 grand last year. Prices have changed, but it would probably still be viable. 15 year solutions aren't easy these days. Lifecycles have diminished with price.

Once again, "it depends"

[thomasinx]

It depends....

Do you care what OS it runs on? (It'll be harder if he wants to keep using windows 95..) For reliability, I'd suggest windows 2000, since it will also work with most recent drivers. The trick will be getting his old software working on it. However once you get the whole setup working, it will be reliable.

How much effort do you want to put into it? You could make this quite reliable by mirroring some 4gb drives, and telling your dad to replace broken ones with spares set aside. Since 4gb drives are pretty cheap, this is a relatively simple solution. (Since 15 years from 1 hard drive is pretty unlikely, use cheap replacements, since space doesn't matter)

Do you care about the speed of the machine? If the only need is to make it keep working, (no real compatibility with existing technologies) this could pretty easily be done with anything in the area of a P3 or P4. These can be pretty cheaply picked up at a lot of used computer stores.

Although, no matter what you do, you're not gonna be able to just buy an off the shelf machine and get this kinda reliability.

Re:Once again, "it depends"

[Scrameustache]

Do you care what OS it runs on? (It'll be harder if he wants to keep using windows 95..) For reliability, I'd suggest windows 2000, since it will also work with most recent drivers.

I'd be worried about Microsoft (or any closed source vendor) dropping support for older OSs, there may very well still be exploitable security bugs in there that could go unpatched.

Re:Once again, "it depends"

[Anonymous Coward]

_Nobody_ provides 15 years of support. That said, his best bet is to go with Microsoft. They are guarenteed to be operating in 15 years. Cant say the same about the other random linux companies.

Also, I dont think anyone can realistically afford to pay Linux kernel dev hourly rate to get them to patch any random outdated kernel branch. Hell a MS Windows license is like 50 bucks when you get it with your PC and you get plenty of years of free updates. Much more than the 'free' versions of (LTS) linux distros without the service contracts.

Re:Once again, "it depends"

[arth1]

Security schmecurity. If this is a system that isn't connected to the network in any way, and where no-one sneakernets in malware from the outside with CDs, floppies, thubmdrives or otherwise, the security you should be concerned about is whether someone will steal it.

In fact, I would recommend against patching any bugs unless they cause a real problem. Stick with what works, and avoid any kind of changes.

Re:Once again, "it depends"

[Narishma]

Does it matter whether it's patched or not if it's not connected to the Internet?

Re:Once again, "it depends"

[Anonymous Coward]

With systems not connected to any public network, that don't exchange any data with anyone/anything except among themselves, you don't have to worry about security too much. If he's using DOS and Win95 since 15 years, there's a pretty strong indication that security is not a top priority.

Re:Once again, "it depends"

[Scrameustache]

Does it matter whether it's patched or not if it's not connected to the Internet?

Someone might still bring in infected files through hard media, and even though he's not going to be a productive zombie, an infection might have dire consequences for the stability of his system. Or it might be one of those viruses that wipe your system.

And if files move in and out of his system (I'm moving out of town, can I have fluffy's records to bring to my new vet?), he might be a sort of digital Typhoid Mary.

Re:Once again, "it depends"

[Pentium100]

Security bugs don't always matter. If the system is not connected to the internet and is not used to open files received from other computers then no virus will be able to get to it. Also, virus creators go with the OS and probably the majority of recent viruses wouldn't work on Windows 95 or NT4 not to mention 3.11.

Re:Once again, "it depends"

[Anonymous Coward]

Do you care what OS it runs on? (It'll be harder if he wants to keep using windows 95..) For reliability, I'd suggest windows 2000, since it will also work with most recent drivers.

I'd be worried about Microsoft (or any closed source vendor) dropping support for older OSs, there may very well still be exploitable security bugs in there that could go unpatched.

Who cares if it is not on the internet?

Re:Once again, "it depends"

[againjj]

From the description, they are not connected to the Internet ("two simple workstations, linked with a network cable"), so security bugs are likely not an issue.

Re:Once again, "it depends"

[Scrameustache]

virus creators go with the OS and probably the majority of recent viruses wouldn't work on Windows 95 or NT4 not to mention 3.11.

Security through obsolescense? :)

Re:Once again, "it depends"

[Scrameustache]

From the description, they are not connected to the Internet ("two simple workstations, linked with a network cable"), so security bugs are likely not an issue.

My worries are not limited to security... there could another "Y2K" lurking in older OSs.
Or even external factors that can mess with hardcoded automation: The government could choose to mess with daylights saving times again, etc.

If it ain't broke don't mean it'll hold up :)

Re:Once again, "it depends"

[Pentium100]

Kind of. Find any rootkit virus that works on Win9x or 3.11 :) All others, while still being harmful, can be easily detected and removed nowadays. Rootkits, though, still are a problem that can make one format the hard drive.

Most 15-year old Sun workstations are still useful

[blind biker]

Take an Ultra 1 or Ultra 2 - they are still rather useful computers, and the OS they run is rock solid by any standards. And there's a ton of software for them.

Now, I'm not suggesting that a 15 year old Sun Ultra 1 would be what your father is looking for, just that it is possible to have hardware that is both good quality and long lasting, and that it would run an OS and software that is still relevant nowadays. Sun did a great job at keeping Solaris backwards-compatible, both hardware-wise (supports older architectures) and software-wise (you can run a lot of vitnage software even on the newest Solaris).

Anyhow... an Ultra 1 is still a damn good computer.

I have little to contribute

[Scrameustache]

But I'd just like to say that this is one of the most interesting "ask slashdot" questions in a long time, and I look forward to replies from my more knowledgeable peers! :D

Ok, small contribution: The dad obviously doesn't need much power, so maybe this would be a good time to make him switch from windows to a bare-bones open source solution which will be most likely to still be supported in 10-15 years, as opposed to the much shorter upgrade-and-obsolescence cycle of Redmond.

Re:I have little to contribute

[troll8901]

I believe people who says they have little to contribute, are sometimes the most reliable people to ask, because they consider more factors, yet are in no hurry to show off their knowledge.

I admire your courage and humility, Mr Scrameustache. I really do.

Re:I have little to contribute

[Moridineas]

I agree that this is an interesting question, and the replies show a lot of interesting points of view. Some more rational than others IMHO :)

You make an interesting point about open source...and I would totally agree with you if they were to keep the software and OS updated. But if we go back in the opensource world 15 years to 1994. Linux 1.0 isn't out yet! Ditto for Apache, no 1.0 release yet. Samba in very early stages of development. PHP doesn't exist yet. Perl is still in version 4! FreeBSD is at version 1.1.

Now, try getting support for any of those versions, or try compiling any of them on a modern OS :)

Re:I have little to contribute

[jopsen]

Honestly, do you really think any distros today will be supported on 15 years?

Seriously, it be hard to even pick a highlevel programming language and framework that won't be obscure or incompatible in 15 years...
If you could just do that, then who cares if there's disk failure in 5 years, and a motherboard in 10, if you can just find a programming language and framework, that'll still run the same stuff in 15 years...

Can you even be sure that x86 won't be somewhat obscure in 15 years?

Re:I have little to contribute

[Scrameustache]

Now, try getting support for any of those versions, or try compiling any of them on a modern OS :)

Ok... I admit won't even try, but I believe there's a few geeks out there maintaining those in their lovely obsessive-compulsive way :)

Re:I have little to contribute

[Scrameustache]

Can you even be sure that x86 won't be somewhat obscure in 15 years?

Now that you mention it, I'm almost sure of the opposite...

The thread is starting to make me think it can't be done.

Re:I have little to contribute

[rubycodez]

but just five years later from ten years ago Linux use started to really blossum: just over ten years ago we have redhat 5.2, I still have the CDs for it and if someone had a Perl 5.005 or C app on it there would be plenty of folk out there who could provide paid support and even mods. So if someone were on THAT I think they would be in good shape for being useful for another five years as a non-internet application platform. Some low-end industrial grade PC from 1998 would be fine hardware platform for it too.

Re:I have little to contribute

[iggymanz]

there are MRP systems for manufacturing plants written in BASIC with commercial ISAM libraries that have moved from platform to platform over the years, from vms and unix and mpe that now run under Linux. My clients have some of those 20+ year old bitches, still going strong on newer hardware, same damn app.

Re:I have little to contribute

[julesh]

Honestly, do you really think any distros today will be supported on 15 years?

Oh, I don't know, if you pick up debian unstable it'll become testing in about 3 years, stable about 4 years after that and maybe reach end of life after another 5. That only leaves you 3 years of unsupported status.

Joking aside...
Seriously, it be hard to even pick a highlevel programming language and framework that won't be obscure or incompatible in 15 years...

If you'd asked me that question 15 years ago, the answer would have been: C++/MFC, or maybe Visual BASIC. I'd have been right with the former and there's a lot of resources for dealing with the latter.

Can you even be sure that x86 won't be somewhat obscure in 15 years?

Yes. Because even if we all upgrade to something better, there'll be a _lot_ of need to run legacy software. 15 years isn't that long in terms of enterprise computing. There will be a solution to running x86 apps.

Re:I have little to contribute

[Anonymous Coward]

...as opposed to the much shorter upgrade-and-obsolescence cycle of Redmond.

I always thought that one of MS's major problems was their long-term support for legacy hardware and software. I think that's only now changing in effort to decrease difficulties in locking down the OS.

Re:I have little to contribute

[Moridineas]

Yes, I don't doubt it at all. At my workplace we use a piece of accounting software that is probably 30 years old. Though it's now a win32 exe, it's still a text console window only, etc. It's commercial and still supported. I would think it likely that the questioner's dad's vet software is similar.

Open source isn't supported that long--that was my main point.

Replacement cost has drastically reduced

[maroberts]

The original systems probably cost $5k-$7k 10-15 years back. Systems to replace these will cost $1-2k and deliver much higher performance. Tell him not to worry about lasting 10 years as the investment cost is not so high. He needs a backup system which it sounds as though he hasn't had. It sounds as though his backup can simply be a couple of USB keys which would hold all his data.

Re:Replacement cost has drastically reduced

[Fenresulven]

I don't think he worries about the cost of upgrading, I think he worries about having to learn how to use the new system. I'd bet that's the reason he wants such a long lifetime.

Re:Replacement cost has drastically reduced

[Javaman59]

I don't think he worries about the cost of upgrading,...

Yep, I'm sure that's what it's about. He's been successfully running his business for 15 years without having to think about his computers. He want's to think about the computers for a short time now, spend whatever money it takes, and then not think about them for another 15 years.

Re:Replacement cost has drastically reduced

[anon mouse-cow-aard]

I'm assuming the person is getting on in years, and so would appreciate a large screen. Small, fanless atom based units can be had, such as: Get a fit-pc (http://www.fit-pc.com) or an eee box (http://hothardware.com/News/ASUS_Eee_Box_B202_Details_Emerge/ ) and can be hooked up to a large monitor (or even attached to the back of the monitor, for zero foot print.) The systems come with SATA HDD, for long life, The only change I might make would be to replace the HDD with an SSD. count on 300 US$ for the server, and 400$ for a nice monitor, kb & mouse. (another 200$ for the SSD)

As others have suggested, put the whole existing system in a vm or emulator...and he will be able to use the same, familiar system, and make it fill the whole screen. This will last as long as anything else. and If you ever do need to replace it, by that time, it will be with a cell-phone...

Re:Replacement cost has drastically reduced

[Anonymous Coward]

Buy a generic box for $500. In a few years get another. If he can not afford this than its not really a business - its called a hobby.

Re:Replacement cost has drastically reduced

[julesh]

The original systems probably cost $5k-$7k 10-15 years back.

Really? I bought a P100 system in 1996 that only set me back about £500 ($900ish at the exchange rates of the day). While it didn't last me 10 years, I used that system as a 24-hour-a-day server well into the 2000s. We tend to misremember when price changes happen. Sure, that kind of price was common in the 80s, but by the 90s they were fairly rare.

Systems to replace these will cost $1-2k and deliver much higher performance.

I really wouldn't recommend spending that much unless you _need_ the higher performance. Cheap components run at low speed (i.e., underclocked CPU and RAM) and redundancy (cheap RAID1 pair of 5400RPM disks) would be the best approach here. Ensure the case has plenty of cooling, and that it doesn't need it. Don't install latest version software as it's more demanding than older versions. Linux and old-style X set up with fvwm and a DOS box sounds like all this guy needs, and such a setup can be run without swap off a ramdisk so the HDD is only used when absolutely necessary. Even on a machine underclocked as low as the motherboard will go this would be a blindingly fast setup.

Re:Replacement cost has drastically reduced

[Hognoxious]

I think he worries about having to learn how to use the new system.

Huh? Does he think keyboards and mice change that much?

Now if it's the software that's the issue, there's all manner of virtual machines & emulators & things that are not emulators; hence it's a non-issue.

Re:Replacement cost has drastically reduced

[maroberts]

My 1990 PC (486DX33, 8MB, 200MB drive) was about £2k and I bought a 386 laptop for a little more than that shortly after. I bought a Pentium P60 box about 1995 which I recall setting me back another £2k. I still use the first keyboard I bought as it is much more tactile than new ones. Incidentally, I was figuring 1-2k as total budget for all the new systems, not just one.

Re:Replacement cost has drastically reduced

[maroberts]

Its possible, but I made no software recommendations. Its obvious his son is planning to configure and possibly maintain it. Making the (possibly) rash assumption that open source packages exist for all his needs, then I would install a Linux distro rather than Windows. Every Windows upgrade often involves a radical change and there is a new one every 4-5 years. Open Source upgrades seem to be more evolutionary and could perhaps be addressed by son doing a yearly upgrade plus one day training on any significant issues.

Re:Replacement cost has drastically reduced

[Amasuriel]

You don't need 1-2k investment. As said at various points already:
1) Make a Win95 VM and load up the crusty old software
2) Buy cheap PC (500 or less)
3) Install VMware player
4) ???
5) Profit!

If the new system is lucky enough to last 10+ years, great. If not, even if it is replaced every 3-4 years it will still be cheap (especially compared to the original) and because of the VM nothing will change for them.

Just make sure you have a backup system for the VM and a copy of the relevant VMware player software.

Re:Replacement cost has drastically reduced

[julesh]

My 1990 PC (486DX33, 8MB, 200MB drive) was about £2k

That was one hell of a high-end system at the time. I bought my first PC at about the same time, and it was a 286 @12MHz, which set me back £600. 45MB MFM hard drive and EGA graphics that were _ever so slightly_ nonstandard, so that moving the mouse around left incorrect colours every now and then.

Car built for 15 years...

[node159]

Would your father ask you to get him a car that lasts 15 years?

I hate to say it, but lasting the designed life span of computer parts (2 years) seems to be a challenge as of late, and buying quality doesn't seem to gain much.

The failure rates now days have been getting a bit long in the tooth.

Re:Car built for 15 years...

[Toonol]

Would your father ask you to get him a car that lasts 15 years?

Many people do drive the same car for over a decade.

However, I bet the father is not concerned about the cost of the hardware; the problem is the cost and failure rate of changing hardware. Every OS upgrade runs a risk of something not working; every time a configuration is touched, something may mysteriously break. Every change means time and effort spent training, changing work procedures.

I think the best bet is generic commodity equipment and a virtual environment. Then it can be copied onto other generic equipment in five years, without any actual change to the product.

Re:Car built for 15 years...

[Anonymous Coward]

???

Is a 15 year-old car abnormal? Mine is currently 15 years old and I have no plans to replace it in the near future. And that's what I expected from it when I bought it. I know several people that have cars that old. (Of course, none of those are American-built cars).

Re:Car built for 15 years...

[Q-Hack!]

Would your father ask you to get him a car that lasts 15 years?

I hate to say it, but lasting the designed life span of computer parts (2 years) seems to be a challenge as of late, and buying quality doesn't seem to gain much.

The failure rates now days have been getting a bit long in the tooth.

Your car analogy doesn't work. Many people get 15-20 years out of a vehicle. I personally drive a 1991 VW Jetta.

Getting computer parts to last is not hard either. Don't buy bleeding edge. Those brand new super high speed drives today do tend to fail, so buy the slower models that came out a couple of years ago. They have the bugs worked out.

Re:Car built for 15 years...

[timeOday]

In fact half of all cars on US roads are at least 10 years old.

Re:Car built for 15 years...

[setagllib]

The slower models were bleeding edge back then too. Apart from the firmware updates nobody applies anyway and make little difference, old drives won't magically become more reliable. If anything I'd prefer newer drives as they likely have higher quality bearings.

Re:Car built for 15 years...

[stonewallred]

Yeah!!!!!!!!!! 5spd and 300K miles and still gets ~30mph around town driving it like a drag car.

Re:Car built for 15 years...

[morari]

What, do you replace your vehicle every other year or something? That must be awfully expensive. I drive a 1973 Volkswagen Beetle (with a fiberglass, kit car body) and have a 1972 Volkswagen Super Beetle as my secondary car. They probably run better than anything you could buy new nowadays.

Re:Car built for 15 years...

[Anonymous Coward]

the failure rates are incredibly low nowaday but the numbers of operations per seconds is so high that a one in a billion chance of an error per operatiom could mean one error per second.....

Re:Car built for 15 years...

[Anonymous Coward]

The car analogy doesn't work, but not because you drive a 1991 VW Jetta.

It doesn't work because computers are at a different stage of engineering evolution compared to cars.

Re:Car built for 15 years...

[Anonymous Coward]

My car's 15 years old. I fully expect to get another 5 of years out of it without any major repairs.

My current computer's over 3 years old too. Most of the computers at my work are over 5 years old and work fine. You're way too pessimistic about how long things actually last.

Re:Car built for 15 years...

[Sj0]

Geez, my truck is 25 years old, and it's great. Not only is it incredibly reliable, but when something does break, parts are cheap as hell because it's all either mechanical or simple electronics.

Re:Car built for 15 years...

[BigGar']

Yes, but how many of those have been in continuous use with no service done to them?

Re:Car built for 15 years...

[Anonymous Coward]

Would your father ask you to get him a car that lasts 15 years?

I hate to say it, but lasting the designed life span of computer parts (2 years) seems to be a challenge as of late, and buying quality doesn't seem to gain much.

The failure rates now days have been getting a bit long in the tooth.

As already posted, tons of cars last that long. I'd expect them to.

I build for clients regularly but my home systems generally last 4-5 years each before I replace them. Even then, it's not that they need to be replaced, just that I have the parts and feel like an upgrade.

Re:Car built for 15 years...

[Anonymous Coward]

Toyota
Honda
Volkswagen

it's not that hard. american car companies are the primary culprit in making 'disposable' cars so there will be a market for next year's model. the american auto industry is only starting to recover from this mentality.

Re:Car built for 15 years...

[CaptainBruce]

Don't confuse the car parked in your driveway with a vehicle used for work. Some company's fleet vehicles are ready for retirement after 2-3 years if the mileage is high enough. That's not always simply because "it was in the lease agreement". The truck I drove at one job had a spectacular transmission failure at 2 years. But it was passed the 250k km mark in their "fix it or auction it" program. It still looked good on the outside. So I think if his father was looking for a >80,000 km/yr car (ie a work vehicle) to last more than 15 years guaranteed, it is an apt analogy.

Re:Car built for 15 years...

[BiAthlon]

No, they don't. You WANT them to run better than what you could buy new. In reality today's cars and trucks are better than just about anything we've ever had. When you combine safety, performance, and reliability they are better than anything we've had.

Re:Car built for 15 years...

[Anonymous Coward]

Volkswagon? The Volkswagons that are 15 years old right now are old enough to be quality vehicles, but the newer ones are falling apart after 5-7 years. They are really the worst of both worlds - the reliability of a domestic, and the repair bills of an import. Except that even the Americans generally have figured out how to make power windows that don't break and how to keep water from leaking into the cabin every time it rains.

Either get a Honda or a Toyota, or get an American made car on a "dated" (read: proven) platform like a F150 or a Mercury Grand Marquis that can be fixed cheaply and parts can be found anywhere. Or take your chances with a Korean car - they are much better than they were a few years ago. But stay clear of Volkswagon.

The current system is flakey

[syousef]

The current system is getting unreliable. If your father leaves it 10-15 years between upgrades the next system will be even less reliable.

Also realize that no matter how much you and your father dislike it, current machines aren't built to last as long as old machines were. The parts can do amazing things but wear out more quickly. I don't know if you'll get 15 years out of a modern disk drive (but then consider that a Gigabyte on one drive would have been a far fetched dream 15 years ago but is commonplace today).

The best thing to do is plan an upgrade cycle every say 5 years. Even then you might need to either buy spare parts of upgrade sooner than you expected if a key component fails and you can't find a spare. One way to combat this is to buy spares in advance but this will end up costing you a lot more in the long run since computer parts get cheaper over time and leaving it to the last minute can save you a fortune. You may also be able to replace older parts with parts that give you new capacity or capability if you adopt a just in time approach.

The other reason to go with a more reasonable upgrade cycle is that computers now tend to be interconnected, and having a 10 or 15 year old system you can no longer patch for security holes means its not as safe to leave on a network of any kind.

In other words, convince your father to upgrade more often (5 years max), backup your data, and buy spares for critical parts but only if you absolutely have to. Unfortunately the pace of change has increased and not putting time into upgrading more incrementally will make the big bang implementation you have to do much riskier.

Re:The current system is flakey

[tomhudson]

The other reason to go with a more reasonable upgrade cycle is that computers now tend to be interconnected, and having a 10 or 15 year old system you can no longer patch for security holes means its not as safe to leave on a network of any kind.

The two machines only talk to each other - nobody's throwing a downloaded game cd or web browser on either of them.

Re:The current system is flakey

[syousef]

The two machines only talk to each other - nobody's throwing a downloaded game cd or web browser on either of them.

Do you ever plan to:
- plug in a USB key
- attach an external drive
- insert a CD

They're all attack vectors. You're not safe just by staying off the Internet.

Re:The current system is flakey

[tomhudson]

USB keys aren't an attack vector on Win95 - no USB support.

Ditto for the external drive - how are you going to get the OS to talk to the non-functioning USB port

CDs? This setup has been used for 15 years as a 2-machine isolated setup. Nobody's going to be inserting suspect CDs into it. Any program trying to auto-install will fail - too many libraries missing/out of date - bad/missing msi installer, etc.

Re:The current system is flakey

[Ash-Fox]

USB keys aren't an attack vector on Win95 - no USB support.

I had USB support working fine on Win95 actually.

Now, if you want USB2.0 support, you want Windows 98.

You don't.

[SatanicPuppy]

15 years ago systems were night and day with the way they are now, and it's only going to get worse. After 10 years you won't be able to find anyone to work on the legacy stuff (unless you buy a proprietary unix system), and there is no guarantee for new parts.

The only way you've gotten away with it is that you have one application which has a very limited required environment, and drive interfaces have only changed once. If you stick with that philosophy, and get lucky with the drives again, you may be able to get by with something similar.

If you have to (which I don't recommend) then pick up a midrange quad core server with a ton of RAM and plenty of room for extra drives. Put a Linux distro on it: no hope of keeping up with Windows security for 15 years, and forget Mac, they're very prone to changing interfaces internally, and then discontinuing the old products.

Then use the server to push whatever app you need to some low duty desktops. You could use a web app, or a client/server desktop app. Again, you're probably good with a *nix.

Your biggest fear is drive space. In 15 years you won't be able to buy the drives you're using today, but there is no point in stockpiling them: they'll be dead in the box after 15 years. Solid state won't fail in the box (probably, but they're too new for it to have been tested) but you may have to replace them more often, depending on your utilization.

Just from personal experience, you're much better off buying a modest new system every 5 years, than a major new system every 15. It's cheaper, and the chance of a catastrophic failure are lower.

Re:You don't.

[Ian Alexander]

If you have to (which I don't recommend) then pick up a midrange quad core server with a ton of RAM and plenty of room for extra drives. Put a Linux distro on it: no hope of keeping up with Windows security for 15 years, and forget Mac, they're very prone to changing interfaces internally, and then discontinuing the old products.

Then use the server to push whatever app you need to some low duty desktops. You could use a web app, or a client/server desktop app. Again, you're probably good with a *nix.

I think that's overkill for one veterinarian.

I suggest going virtual on commodity hardware that changes out every half-decade or so. I suspect that as long as your virtual machine itself doesn't change the cost of buying hardware with the chops to run it will only decline over the years.

Re:You don't.

[644bd346996]

Clearly, this guy doesn't even need internet access, so there's no need to worry about operating system updates. All that's needed is reliable hardware and future-proof communication ports for backing up and migrating to the next system.

A low-power fanless ARM or x86 based system with a SLC flash drive and an ethernet port is all that's needed to meet those specs. The longevity of the flash drive isn't a problem, since even the smallest drives are many times larger than what he's already got, so the wear-leveling will keep the drive going for a very long time.

Re:You don't.

[syousef]

If you have to (which I don't recommend) then pick up a midrange quad core server with a ton of RAM and plenty of room for extra drives. Put a Linux distro on it: no hope of keeping up with Windows security for 15 years, and forget Mac, they're very prone to changing interfaces internally, and then discontinuing the old products.

How many of your machines are running upgrades rather than reinstalls of Linux distros taht are 15 years old.

Three examples of sufficiently old distros:

Redhat/Fedora - You'd have literally had to start with Readhat version 1.0 - http://en.wikipedia.org/wiki/Red_Hat_Linux#Version_history - and have moved to Fedora 9 or 10.

Debian - You'd have to have started with 0.9 - http://en.wikipedia.org/wiki/Debian#1993.E2.80.932000

Slackware - Looks like you had to start with verison 1 or 2 - http://en.wikipedia.org/wiki/Slackware#History

It's nice in theory but has ANYONE EVER done this???

Re:You don't.

[cyrano.mac]

Macs don't change interfaces all that often... I've got a 10 year old G4, 400 MHz, 768 MB RAM, 20 GB HD, running OSX 10.4.11, with IDE, USB, Firewire and SCSI. It's got a weird bunch of disk drives attached: CD/DVD, DVD-RAM, Syquest, Bernouilli, MO, Zip. Both tapes have died a couple of years ago and since I rarely used those anyway, I haven't replaced them. It has no floppy, tho ;-) I use it for data transfer and recovery whenever one of my clients comes up with old stuff from their archives. All the hardware is old. The Syquests are approximately 12 years old. But the recovery software is the latest version.

Re:You don't.

[iggymanz]

eh, I can buy the seagate drive used in many 1987 AT compatible for $100, with warranty. Reason is that there are plenty of industrial and voicemail systems (often with OS/2, w0h00) that use that bitch. Installed a couple myself for clients. Funny one can still buy new 80486 computers too, industrial apps again.

Re:You don't.

[AdamHaun]

15 years ago systems were night and day with the way they are now, and it's only going to get worse. After 10 years you won't be able to find anyone to work on the legacy stuff (unless you buy a proprietary unix system), and there is no guarantee for new parts.

What do you mean it's only going to get worse? The rate at which people upgrade their hardware has been slowing down, not speeding up. You might not be able to find replacement parts in ten years except in bargain bins and eBay, but that's no different from now. As for people, there are plenty of people who remember how to work on Windows 95 and there will be plenty of people in ten years who remember how to work on Windows XP.

Furthermore, your hardware recommendation is complete overkill. The guy is running a *DOS application* on a *486*. Even the cheapest system you can buy today would be at least twenty times faster, and the program and all its fifteen years of data could fit in a 512MB stick of RAM. Why on earth would he need a quad-core server? The program he needs to run is single-tasking! It sounds like he doesn't even have internet access. He doesn't need a new architecture. He probably wouldn't even need a new computer if Windows 95 weren't so flaky to begin with. Unless he's planning a major software upgrade (which it sounds like he isn't) there's just no way that his computational demands are going to grow enough to need what you're suggesting.

Re:You don't.

[Anonymous Coward]

If you have to (which I don't recommend) then pick up a midrange quad core server with a ton of RAM and plenty of room for extra drives. Put a Linux distro on it: no hope of keeping up with Windows security for 15 years, and forget Mac, they're very prone to changing interfaces internally, and then discontinuing the old products.

Then use the server to push whatever app you need to some low duty desktops. You could use a web app, or a client/server desktop app. Again, you're probably good with a *nix.

I think that's overkill for one veterinarian.

I suggest going virtual on commodity hardware that changes out every half-decade or so. I suspect that as long as your virtual machine itself doesn't change the cost of buying hardware with the chops to run it will only decline over the years.

Actually, I'd recommend doing exactly what the GP said - except dual-core rig from eBay should be fine. Get a unit that has at least dual power supplies, ability to hold at least two drives (preferably 4) and has a couple of dual-core CPU's. Or a 2-4 CPU P4 server. Ebay's got em under $600 for the main machine, and if you're clever, you could buy a couple old P4 servers (of the same model) and glom the parts together, while simultaneously getting a spare motherboard, power supplies and chassis. Mix n match the RAM and CPU's to max it out as much as you can. Use CentOS with Xen to get all the virtual goodness you'd need and let it fly. Hopefully a real server will be able to self-monitor and email you if something dies (which will probably be the UPS first).

The big advantage of this over commodity hardware is that old server tech gets cheap quick, and dual power supplies buys you some time if one fails, so that he's still in business while you swap it out. Commodity HW is usually single power supply, single fan, and that risks unscheduled downtime.

Re:You don't.

[jbeach]

This guy's using Windows 95 to run his business as a veterinarian. I expect he doesn't need or have the time or interest to learn linux. Drive space is no way an issue, either; the guy probably doesn't even listen to mp3's...

It's fine to love *nix for it's many values, but most people don't need them and the learning curve is steep. Most people just need something they're familiar with to get them through their day.

Short Answer: Windows Server

[yakatz]

In one of my offices (programming POS and web applications for a retail store), we have been running happily on two Windows 2000 Servers (one is SBS) for over 8 years.

Long answer: some of our network closet is held together with duck-tape.

Re:Short Answer: Windows Server

[Anonymous Coward]

I think his problem is hardware related. Surprisingly there are still computers running win95 around.

Re:Short Answer: Windows Server

[Anonymous Coward]

They make tape out of ducks?

Re:Short Answer: Windows Server

[__aaclcg7560]

Plus a few machines still running Windows 3.11 and DOS. I did one-night tech job a few years ago converting a Token Ring network to Ethernet network. I seriously never expected to see a full blown Token Ring network in the field.

build your own and amd is more backwards-compatibl

[Joe The Dragon]

Just use standard parts and you should be able to easy replace stuff if it brakes down with a standard part no need to replace the same video card, psu, hdd or other part just replace it with a new one that uses the same bus.

Re:Build your own system. Keep it solid and simple

[betterunixthanunix]

"As long as you buy standard components, they should be easy enough to replace 15 years from now, don't go hog-wild stocking up on replacement parts"

You'd be surprised. I have a 15 year old desktop that takes ISA cards, and I have seen younger systems (relatively speaking) that only took AT keyboards. What seems like a standard technology now, that will "never be replaced," may very well be long forgotten in 15 years. 15 years is a long time to try to keep a single system operational; I would suggest an approach based on virtualization, so that if the entire system needs replacing in a few years (e.g. a mobo failure with no compatible mobo available), the migration process is not so painful. Also, use some sort of RAID, but be careful as it may not be possible to find a compatible replacement hard drive in 10 years -- again, virtualization may solve this problem.

if he's ok with DOS, just buy a new machine

[sneakyimp]

You could get computers about 15 times more powerful than his old system with 1000 times as much storage for under $300. My guess is that the hard drives are definitely going to fail in 10-15 years -- possibly several times. You could get a motherboard with RAID 5 and that would help prevent data loss.

You would also need an install disk for Windows 95 or Windows 98 -- try to get the most recent operating system that this dinosaur can still run on.

I would also expect that the fans and the power supply might get kind of tired after 15 years.Another consideration is the connection technology of power supplies and hard drives. Hard drives are all SATA now. I'd bet that will change at least twice in 10-15 years so you might consider buying a few extra hard drives -- but only when the next big hard drive connect technology has been announced. They'll be cheaper then.

If you want to replace the software, you *might* consider migrating his data to the cloud or something (google docs or Amazon EC2 or something? I don't know what his software does). Changing his software setup is a lot more work though and you could incur significant time and effort and expense extracting the data from the legacy system and getting it into some new format. On the other hand, if you trust Google/Amazon/whoever to be your cloud provider for the next 10-15 years, you don't really have to worry about the machine.

Re:if he's ok with DOS, just buy a new machine

[Anonymous Coward]

You could get a motherboard with RAID 5 and that would help prevent data loss.

The fallacy of RAID-5 continues.

Once you lose a single disc in a RAID-5 the odds drastically rise that you'll lose another while rebuilding the array, losing ALL of your data in the process.

RAID-1 or RAID-6 are much more reliable choices, but in this case neither is really warranted. Just use a couple USBs or an external drive for backup.

Re:if he's ok with DOS, just buy a new machine

[NormalVisual]

On the other hand, if you trust Google/Amazon/whoever to be your cloud provider for the next 10-15 years, you don't really have to worry about the machine.

You do have to worry quite a bit more about connectivity though. It sucks to have your business shut down for a couple of days because some idiot with a backhoe down the street wasn't paying attention.

Re:if he's ok with DOS, just buy a new machine

[tomhudson]

Just use a couple USBs

Win95 == no USB support.

I'd say just clone the system to a new hard drive (partitioned to 4 partitions of 2 gigs or less, and just ignore the rest of the space), shove the hard drive in a new box, with not more than 256 megs of ram, and if the old software and OS runs (which it might - it's old, so it's not exactly bleeding-edge), problem solved. Alternatively, there's some open-source animal shelter software that might have enough included functionality for veterinarieans - after all, shelters do have vets either on call or on staff, and they do do billing for things like fixing the animals, shots, sales of animals, food, toys, accessories, putting them down, etc.

All this talk of quad core / raid is just stupid. Today's netbooks are "good enough". Heck, a smartphone would probably be good enough if you could hook it up to the printer.

Re:if he's ok with DOS, just buy a new machine

[hawkinspeter]

Never use RAID5!

He'd be much better off with full mirroring (RAID1) done with software RAID so that he can just use standard drives and connections. Using software RAID means that can you survive a motherboard failure by simply dropping in a new one that can talk to the drives and away you go. Otherwise, you'd be tied to the chipset used for the RAID.

Also, RAID5 requires at least 3 disks, so you're actually increasing the chance of hardware failure compared to a 2 disk RAID1 (plus RAID5 sucks when a drive fails and puts additional stress on the remaining drives during a lengthy rebuild process).

Re:if he's ok with DOS, just buy a new machine

[Anonymous Coward]

Wrong. if you have only a single disk, a single disk failure definitely loses all your data. With RAID5, a single disk failure might lose your data but only if you aren't quick enough in replacing the failed unit. Not perfect, but a big improvement.

Re:if he's ok with DOS, just buy a new machine

[maxume]

Please explain how RAID1 puts less stress on the remaining drive in the case of a failure.

Re:if he's ok with DOS, just buy a new machine

[Ash-Fox]

Win95 == no USB support.

I had USB devices working fine in Win95. Maybe you should update your copy of Windows 95? Updates are still available on http://windowsupdate.microsoft.com/ for it.

points of failure

[Tumbleweed]

Okay, here's my thinking: consider what things fail on a system, and why.

Number one thing that shortens system life of your average non-overclocked machine: bad power. This includes a crappy PSU, and bad power coming into the machine from the wall. The solutions: PC Power & Cooling PSU, and a good power-regulating UPS. Keep in mind if you have some severe power surges, you may have to replace parts of your UPS over the years. Better parts in the UPS than in the computer itself. It's there to take the hits; accept it's brave sacrifice and move on.

Number two: cooling. Cool your system! Make sure the system shuts itself down when fans stop running, so the system doesn't fry itself. Replace your fans every 2 years. They're cheap ($10 per fan for the good ones). Don't forget the fans on the HSF (heat sink fan unit on the CPU) and on your video card. Or better yet, get a system with integrated video. One less fan to worry about, and if it dies but everything else on the mobo works, you can always add a videocard later.

Number two A: underclock the system a bit. Nothing drastic, but why push it, since you're going for longevity?

Number two B: Keep the system cleaned out on a regular basis. I'm not talking about defragging (we'll get to that), but keep the insides free from actual dust. Regularly. That doesn't mean once a year, but at least a couple of times a year. More often depending on the environment it's running in. Some cases have filters where air is bring drawn in from the fans, but almost all cases will be bringing in air from more places than just the fan intakes, so you'll get dust no matter what.

Number three: crappy RAM. Look, just buy quality components, run Memtest86 whatever on it to make sure it's good when you get it. Again, don't overclock this, or anything in the system.

Number four: quality mobo. This includes things like solid capacitors (not 'solid-state', which some people confuse this with), etc. Read the reviews from the hardware sites and make your choice. This kind of thing is where you'll most likely have to build your own machine. It's not that hard, so don't worry about it if you've never done it before.

Number five: storage. Okay, here's something that's in flux right now - spinning drives or SSD (solid state disk, meaning no moving parts). I'd say go with an SSD now, and upgrade in 2 years or so once things have settled down a bit. Higher-quality SSDs are already pretty nice, especially with wear-leveling, etc. Make sure you dispose of these properly, as the ability to reformat an SSD isn't as secure (as far as I know) as with spinning discs, yet, to a level I'm comfortable with. When an SSD dies, the information is still readable, generally, so you can't (maybe) reformat it completely. The bit gets marked as bad (non-writable), and I'm sure someone could (or has already) written a program that can read those bits to get information that you would hope is gone after a format. The rise of SSD for storage will be one of the biggest boons to long-life systems, as long as you follow the above advice first.

Number six: If he needs a floppy drive for some reason, turn off the auto-check feature on it. That kills floppy drives faster than anything. I've had floppy drives last longer than ten years. So long that the entire format went obsolete before the drive itself died.

Number seven: security. Put the thing behind a firewall or at least a NAT 'firewall'. Lock the machine the hell down; no root/administrator access to normal users, etc. If in Windows land, install Vista/Win 7 - the sandboxing abilities are far superior to that of Windows XP & older. Obviously, some type of Linux OS is preferrable. Security is a process, remember, so you have to keep checking things out.

And that's all I can think of off the top of my head right now. I'm sure there will be plenty of other good advice to come along from others.

Re:points of failure

[Renraku]

Actually, building a barebones system shouldn't be this complex at all.

You can build a decent machine that doesn't require a fan on the CPU at all. Granted, it wouldn't be very fast, I don't think the current setup is very fast either.

Just use onboard sound/video/data and use a cheap 200GB or so drive that will cost you $30 to replace, new, if it dies. Backup regularly and it shouldn't cost you more than $200 or so to replace most of the system. A UPS is important, though, as it will greatly extend the life of your power supply if it gets out of the infant mortality stage.

Re:points of failure

[LWATCDR]

Will Vista Win7 run his old dos program?
Unless new software is part of the plan that is important.
I agree with most of your suggestions but would add one.
You can build a fan less PC. If you passivly cool everything then you will also cut down on the dust issues.
SSDs sound good. Maybe a RAID 1 with a hot standby would be a good plan as well.
I would suggest Linux with Virtual Box or DosBox as an option if they want to keep the old software.
I would then keep a golden backup of the OS image so that you can do a rapid restore of the OS when it gets junked up.
Also he may have NO need for a firewall at all. He may not have those systems hooked to the internet. Not every system must be hooked up to the internet.

Re:points of failure

[maxume]

I would think that if you wrote 0s to every block of an SSD, that there wouldn't be anything recoverable left on the SSD (I am also probably naive enough to believe this about HDDs). Are there rumors of tools that can tell the firmware to ignore remapping and such?

Re:points of failure

[Tumbleweed]

I would think that if you wrote 0s to every block of an SSD, that there wouldn't be anything recoverable left on the SSD (I am also probably naive enough to believe this about HDDs). Are there rumors of tools that can tell the firmware to ignore remapping and such?

This is where SSDs differ from HDs. When a block (or cell, or whatever it's called) on an SSD dies, I've read it tends to do so on write, thus remaining readable. So once it's been remapped, PHYSICALLY, it can still be read from by something that can ignore remapping - I'm quite sure the three-letter agencies have already got something like this, and obviously the SSD makers will already have this. Data recovery companies will likely, as well. For purposes of entities who have the legal requirement of securely destroying data (credit card companies, those with access to medical records, etc.), you can no longer use a multi-write format like you describe, as the media isn't physically capable of writing those 0s to all bits that are still readable. I dunno what a degausser would do to an SSD, but I'd be interested in finding out. We have one at work for our HDs. You wouldn't believe how loud the thing is.

Re:points of failure

[maxume]

Yeah, okay, I was aware of the die-readable behavior but completely failed to consider it, thanks for the quick explanation.

I guess the upside is that storage is cheap enough that by the time it is decommissioned, destroying it doesn't destroy all that much value (at least, in comparison to the market for new storage, there is still waste).

Re:points of failure

[Ash-Fox]

Will Vista Win7 run his old dos program?

Windows Vista and higher does not support running DOS applications natively.

I would suggest Linux with Virtual Box or DosBox as an option if they want to keep the old software.

Which Linux? I don't know of any distribution that will make a version that lasts 15 years of updates. The only operating system I think of that will do that is Solaris.

Re:points of failure

[tapanitarvainen]

you may have to replace parts of your UPS over the years.

You will have to replace UPS batteries several times over 15 years. If you manage to find an UPS that'll have spares available for that long.

Number two: cooling. Cool your system! Make sure the system shuts itself down when fans stop running, so the system doesn't fry itself. Replace your fans every 2 years.

Or go for passive cooling. Make sure it's efficient enough, though: running too hot will also reduce component lifetime

keep the insides free from actual dust.

There are passively cooled cases with no air intake at all. They can't handle heavy-duty hardware, but I understand it's not an issue here, something like an Atom would be enough.

Number seven: security. Put the thing behind a firewall or at least a NAT 'firewall'.

Or don't connect it to Internet at all. His old machine presumably wasn't, so maybe it isn't needed. In any case it's the only way to avoid frequent software updates. Perhaps get another machine, a netbook, say, for browsing the net, pointing out that Internet will change in 15 years so much that any machine used for it will become obsolete much sooner anyway.

Re:points of failure

[Narishma]

Why would he need any updates? From the question, it seems the guy only uses 2 DOS applications and the computer isn't connected to the Internet. If he plans to not use any other software, something like FreeDOS makes more sense than Windows or even Linux.

Re:points of failure

[LWATCDR]

Solaris might be a good option if Sun lives. IBM is not going to buy it and I hear that HP is interested. After what that company did to DEC I don't have high hopes for Solaris if HP buys Sun.
Cisco might be a good match.
I would say maybe CentOS or Ubuntu LTS. The Ubuntu upgrades tend to be pretty pain free so far. Or if he doesn't go on line he may not need updates at all.

Re:points of failure

[LWATCDR]

Well he does have two PC networked running Windows 9x so Freedos might be a step down in functionality.
Using Linux with DosBox or with VirtualBox would allow him some extra features like. Backing up to a USB device, Using Open Office or SQL Ledger. And maybe getting on line and using email and what not.
OpenOffice could be used to do letters and other reminders if he really wants to.
Plus Linux using Dos Box/VirtualBox would allow him to run both Dos programs at the same time. Just like he probably did with Windows 9x.

Re:points of failure

[Ash-Fox]

From the question, it seems the guy only uses 2 DOS applications and the computer isn't connected to the Internet

For all I know, he's using the computers for e-mail and browsing too.

Veterinarian

[robably]

I wouldn't worry about the system having to last for 15 years if he's already a veterinarian. What is that, 140 years old? Wow.

Re:Veterinarian

[RobDollar]

Depends if your counting in dog years.

Re:Veterinarian

[joeme1]

No, it means he only eats veterans. Duh.

Hard disks are your bane

[Anonymous Coward]

Moving parts (as mentioned earlier) --fans, hard disks will die. Build the box yourself. You can get much better components (motherboard, etc) at a lower price. The low price is a bonus, the better components are the real win. Build with as new a kind of technology as possible. Old tech is dying tech. You will likely have to replace hard disks so buying a new style (sata, not ata) means not having to find an ata drive in 10 years (there may be none). More memory is better than less. Also, it shouldn't be hard to back everything (the whole disk) to another disk (or two) in the same box. If one dies, you only lost maybe 1 days worth of data. Don't overclock, pay attention to heat. Cold is your friend here. An incandescent light bulb running at half power will last 50 years. At full power: 20000 hours max. Old computers did not get that hot. New ones do. I built a corei7 box a few months ago. It idles along at 30 degrees C (86 F). When it gets hot, its at 57 degrees C (134.6 F). This is after an aftermarket cooler went in, which dropped the temperature by 10 degrees C! An atom processor will give more performance than he has now, and if cool last for what you are looking for.

Simple answer, is it's time to redesign the system

[ducomputergeek]

I went through this with a company 3 years ago that was running their billing and inventory system off DOS and it still worked with all their venders/payment company. The owners were getting ready to sell and retire (both in their mid 60's.) Contingency of the sale was the upgrade of their systems.

Fortunately, their backend wrote to CSV files. When it came time to choose a new billing system, we found one that ran over generic ODBC and could support a number of database venders including PostgreSQL (which is what we used) and coded the front end in Java.

The software now is no longer dependent on hardware platform. Now moving forward, if a server fails, load the last back up on new hardware and go.

RAID 1

[thue]

We have an old server in the basement with a very custom setup, which I would rather not mess with. It is internet gateway for 300 people, so I would rather not take out of production to play with it.

It has a 3-way RAID 1 (linux software raid), so it will take 3 disk failures for the storage to die. In fact, one of the disks have failed already, but since it still have 2-way RAID it I see no reason to do anything about it :).

The point: Using 3-way RAID is a good insurance for a long lifespan without maintainance :).

Re:RAID 1

[hawkinspeter]

Quick question, why aren't you replacing the broken disk if the machine is important enough to warrant a 3-way RAID1?

Re:RAID 1

[sasami]

It has a 3-way RAID 1 (linux software raid), so it will take 3 disk failures for the storage to die. In fact, one of the disks have failed already, but since it still have 2-way RAID it I see no reason to do anything about it :).

Here are two reasons. If you wait for the next disk to die, then:

1. You have no protection against the final disk failing during the rebuild.
2. You have no protection against bad sectors discovered during the rebuild.

BTW, both of those reasons are why our customers have been demanding RAID6.

Re:RAID 1

[thue]

I am maintaining the machine as unpaid work for my Kollegium. When it fails I will probably just replace that complex system with a quick simple gateway in an hour or two. My current course of action is designed to minimize the effort I put in until I actually have to do something :).

Internet access?

[st0rmshad0w]

Quick question, is either of these systems currently or in the future going to be connected to the outside world?

If so consider this, you have obsolete, out of support software running without anyone developing security patches for it anymore, which is storing client and financial info. The only thing protecting it is its own obsolecense. That is simply not acceptable.

If its an air-gapped network that will never touch the outside world run whatever you like.

My personal suggestion, virtualize the workstations on new hardware, easy to shift from hardware to hardware and maintain the obsolete software and if needed you could create seperate VMs if there is a need for external internet access.

Or if you like you could set up a server and use something like Wyse thin clients on the workstation end, those will probably last you 15 years hardware-wise. The server tho, hard to say.

Basically little to no commercial PC hardware is built to last 15 years.

Better solution? Don't be afraid of upgrading.

[darpo]

Don't count the current systems' 15 year lifetime as a population-wide trend; count it as a fluke. It's probably sheer, blind luck that's enabled them to last that long.

Modern PCs aren't appliances. They are not an oven with reliable, decades-old technology in them. They are if you're NASA, but not in the real world. They're ever-changing. If your business relies on them, you should learn about how they work and not be afraid of them.

Re:Better solution? Don't be afraid of upgrading.

[LWATCDR]

I can see your point but his business doesn't rely on them. He could do everything with paper and pen. The computer saves him a little and time and money. His current software might do everything that he wants to do. For example take a look at the new ATM machines. They can put up pretty pictures now and in full color. Do they do anything more for me than the old ones? Not really. Now the ones that can read the checks you deposit are really cool but the new ones that just have the new pretty UI bring no extra value to me.
An 80 year old hammer is as good as a brand new hammer.

Re:Better solution? Don't be afraid of upgrading.

[try_anything]

I bet the fear comes from the software side, not the hardware side. How easy is it to install the software? Does he know where the install disk for each piece of software are and whether they still work? How hard was it to get the software configured correctly in the first place? Can the configuration be migrated, or will they have to recreate it using trial and error? Does the database make it easy to migrate data from one database to another, or does that require a proprietary tool from a now-defunct company?

Heh, what are the odds that someone who hasn't experienced a failure in fifteen years even thinks about these issues? Or bothers to back up his data?

Re:Better solution? Don't be afraid of upgrading.

[Fortunato_NC]

Not to be pedantic or anything, but do you actually use hammers on a fairly regular basis? An 80-year old hammer would likely have a heavy wooden handle that transmits all of the shock and vibration right into the user's hand. It's likely worn smooth on the face so that the head might wobble on impact, sometimes slipping off the nailhead, causing you to have to re-drive more nails.

Modern hammers are much lighter (have fiberglass handles or are cast as a single piece of metal with a rubber handle), don't transmit as much shock to the user, and have a textured face that "bites" the nail and ensure that the head doesn't slip off when your strike is slightly off-center.

If I actually had to drive nails all day, I'd much rather have a newer hammer.

Re:Better solution? Don't be afraid of upgrading.

[john82]

If he seriously wants a record system from today to last 15 years, the best thing going is pen and paper. Keep two ledger books, but store them separately. Computers aren't built like they used to be. Today they're EXPECTED to fail after 3-5 years.

Re:Better solution? Don't be afraid of upgrading.

[nelsonal]

Anyone who drives nails all day switched to a pneumatic nail gun years ago. Faster and better. They're not expensive, and they'll save you a fortune in labor costs. The hammer is for pulling nails and the occasional nail in an odd location (so the handle doesn't matter nearly as much). Just for the record, I prefer the feel of hickory to rubber when I'm hammering all day.

Upgrade the current workstation.

[Anonymous Coward]

Do not do anything more than strictly necessary. Go on ebay or ask some friends for a 5GB disk and some old ram dimm. Those who don't like to use recent technologies, do not deserve them.

Re:Upgrade the current workstation.

[CAIMLAS]

Those who don't like to use recent technologies, do not deserve them.

They likely don't know what they like. They like what they're comfortable with, more than likely, and don't want to spend a lot of money: they likely figure that new = more expensive, because hey, it's a lot faster and more capable, right?

Also, systems of the vintage we're talking about don't use DIMMs. They used SIMMs. I have not seen a SIMM in maybe 5 years; the last time I saw them, they were very expensive when in working condition. I saw more used for jewelry - keychains and bag zipper pulls, mainly - than anything else.

Virtualization

[n0dna]

This thread was finished on the third post.

Keep it powered on

[donstenk]

Get a UPS, you don't need a powerful machine so get something quiet with low energy consumption and let it switched on at all times.

Notice how a bulb always fails when you switch it on? It is much the same with most electronics - just keep it running!

Why Not Just Update?

[flnca]

If the hardware is still running, there's technically no need to purchase new computers. You could replace the 500 MB drives with up to 4 GB ones (Win95 can handle that much). Do an inventory of all applications installed on the machines, and reinstall them after setting up fresh Windows 95. Or use a tool that can copy a smaller drive to a larger one. If the software works and everything, it should be best left as is. A new system might offer more speed and storage space, and reliability, but the custom application might not run anymore. There are solutions like virtualization, but you could also use Linux running DOSBox (which are both free). DOS applications are very picky when it comes to their runtime environment, so you should test your solution with your father's programs and data on a separate machine first. For durability, I have no idea what to recommend you. PC hardware has become a bit less reliable in the past years. You might try a PowerPC solution (which requires less powerful hardware), and run OpenBSD or something on it. DOSBox or QEmu might be suitable for running DOS apps on BSD. OpenBSD has the advantage that it never changes until you change it, there's no update ever until you do it manually. Another solution might be FreeBSD, which should also work very reliably. For longevity, I would abstain from all things Windows. Recent Windows platforms like XP or Vista that require activation may fail to be reactivated when Microsoft switches off its servers. Also, Windows Updates for a particular platform might be no longer available at some point. So, using a UNIX-like system might be the best idea. If you use Linux, switch off the automatic update feature to avoid breaking the system by update (does happen sometimes). I would use OpenBSD or FreeBSD for a system that needs to last for a long time.

Virtualization is a good answer here

[Kjella]

And what about virtualization options? That lets you move around to various hardware without causing an upset on the system build.

I'd second this. I've had pretty much every component in a computer fail on me, CPU, mobo, graphics card, HDD, PSU, whatever. Shit happens and trying to make sure you're covered with spares on everything in 10 years is highly impractical. Things like a bad PSU can kill a component twice before you realize the true problem or fail twice just for the hell of it Of course you can pick quality hardware, run a pair of disks in RAID1 and all that but having an easy migration path that'll be up an running in minutes on a new box is much better.

However, is he still planning to keep this an offline setup? If so he doesn't have to worry much about security problems, because that's the most normal reason you don't try to do this. Even the most extended support from Microsoft, Apple or Linux distros don't last 15 years. For any internet connected PC I'd certainly want the host OS to stay in support, then run any legacy apps from inside a firewalled virtual image. Seems the easiest way to not mess with what works while keeping the shields up.

Re:Most 15-year old Sun workstations are still use

[Athens101]

I second this idea of old Sun hardware. My old U5 precached six years ago is still going strong.

Re:Most 15-year old Sun workstations are still use

[SatanicPuppy]

Sun hardware is awesome. But so is the price tag, and, if you're not willing to pay support, so is the bill if something breaks.

If you use commodity hardware with OpenSolaris or Linux, you can get some of the same benefits, without the cost.

I wouldn't recommend Sun for a small shop with indifferent data storage requirements. It's worth the money if you're dealing with a lot of money, but otherwise cheap works just as well.

You're overthinking it

[Anonymous Coward]

Take a step back, and ask what's the reasoning for the 10-15 year goal? Cost savings? Having to avoid the effort and time in transition to a new machine?

As others have pointed out, you can't guarantee that long of a life span. You can buy high quality parts (or high quality machines) to make it more likely, but buying lots of spare parts or an entire back-up computer should require a significant justification. Buying spare parts means you're paying for something you don't need now that will cost much less later on.

Here's a thought for a 10-15 year plan: Buy a computer. If it breaks while new compatible parts are being made, buy the necessary replacements. If it breaks and no new compatible parts are made, buy a used machine from around the same time period. If you want some extra insurance and peace of mind: once the cost of your computer becomes low enough - say $50 - buy a compatible machine as a back-up, rather than waiting for something to break.

Dont overthink this

[Anonymous Coward]

Get two new cheap workstations, and get a low to mid-range server machine with either RAID or a fail-over. Check the system health regularly, and monitor for when repair or maintenance is needed. the most expensive thing is going to be the server, but in reality it can be a third workstation that is just dedicated for the database, and email server (if they use one). I set up an ERP on a fairly powerful computer that ran about $700 that had 15 workstations communicating with it, and the only problem was dust build up.

Go server hardware

[Bigjeff5]

Get server hardware. It's the only stuff built these days with reliability as the #1 concern. And get GOOD server hardware. That doesn't mean dual quads with 64gb ram, that means a well known line in a company known for servers. I'd probably go HP or IBM, and for what your father needs you can pick the bare minimum and it will be fine for years.

Remember when you spec this out, that #1 failures are those with moving parts, as others have said already. This means, when you build your server, you want the LOWEST capacity and LOWEST speed you can get, for reliability. The high capacity, high speed drives fail the quickest because they push the hardest. SSD might be a good alternative, but as yet the long-term reliability is unproven and they have a definite limited life-span (i.e. # of writes, how quickly that is used depends on the application), instead of a constant potential failure rate. The plus on that is there should be very little chance of a SSD failing until it actually reaches its end of life.

So, slowest fans you can get, or no fans if possible, and slowest HDD. You should probably go with as low a power CPU as possible also, to keep from taxing the PSU.

Also note, VM would be a heck of a lot of work to get going, but new migrations and failure recovery should be simpler. Gotta pick what works for you.

Re:Go server hardware

[Anonymous Coward]

I forget my slashdot username.

Anyway, I personally like HP Proliants, unlike everything else HP makes, they're built like tanks. They aquired their server line from Compaq - and I have a Compaq Proliant that's a good 10 years or so old (approximately - when did the Pentium II come out?).

Note that HP's consumer computers (laptops and such) tend to be flimsy with non-essential parts that break a lot.

Re:Go server hardware

[Anonymous Coward]

Do not skip the fans. Build a computer that can survive without the fans, but make sure to have sufficient air movement to keep things cool. Most fanless setups are rather hot, which will decrease the expected lifetime of the system. A VIA Nano should be more than fast enough to run the old setup fully virtualized faster than the original. If you don't like those, an Intel Atom is a good alternative, as long as you get a low power chipset to go along with it (the desktop boards usually require a fan for the chipset).

Re:Go server hardware

[Anonymous Coward]

I like the VM ideas. The first thing I thought of though was a terminal server type setup. LTSP or something similar. He could keep his old PC's as clients and just buy a new server ever 5 to 8 years. Could go longer if he feels like risking it though too.

Re:15 years?

[donstenk]

And add to the high tech junk pile just for the sake of it? There are actually still people in the world that:

1) buy only what they need or seriously want
2) spend on quality
3) keep it as long as it works

It is against the consumeristic trends, but really - why on earth would you need a new computer every 3 years to store clients records? To other professions a computer is just a tool like any other.

Actually - why does the 15 year old PC need changing? Just because they are getting a bit long in the tooth - that's no reason at all! (Actually fair enough - and well done).

Or at least

[Colin Smith]

forever

Until the virtualization software is made obsolete by the vendor.
 

Re:Or at least

[darthflo]

"the" virtualization software? Just be sure to keep the vm's disk an actual partition and you can swap your virtualization software in no time.
Alternatively, go the matryoshka way. Run Win95 in (for Example) VMWare 5 on current Ubuntu now, wrap that in Xen on Ubuntu 12.4 LTS, wrap that in the 2016 Edition of Virtualooz on vanilla Lunix 28.6.19 and that in some deep fried beer batter. Processor speed will keep up.

Re:Or at least

[similar_name]

Then he'll just have to virtualize the older virtualization software in the newer virtualization software.

Re:Or at least

[frieko]

I know, what if they release a new version of Windows 95 and you need a software update to support it!!

Re:Or at least

[Anthony_Cargile]

Then he'll just have to virtualize the older virtualization software in the newer virtualization software.

Why does this sound like a version of the "sup dawg" meme, a la this?

Re:Or at least

[setagllib]

KVM is in Linux mainline, so it'll be supported for as long as computers still exist. It's not even tied to a specific CPU architecture. Even if KVM somehow failed, QEMU is more than fast enough to virtualise the machine in question. Original poster can get a cheap desktop and a free Linux install and just migrate the virtual machine if and when the desktop fails.

Re:Or at least

[Anonymous Coward]

I guess when KVM is removed from the Linux kernel?

Re:Or at least

[MasterOfDisaster]

Sup dawg, we heard you liked legacy applications so we put an emulator in your visualization so you can compute while u compute!

Re:Or at least

[Oloryn]

That type of operation actually used to be fairly common on mainframes. The next bigger model machine would come with an emulator for the previous model, and instead of, say, rewriting the payroll application, you run the old one under the emulator. Several mainframe purchases later, you find that you're still running the original payroll program under 6 layers of emulators.

Re:Or at least

[Anonymous Coward]

forever

Until the virtualization software is made obsolete by the vendor.

VirtualBox is open source.

Re:Or at least

[DrgnDancer]

Nothing is EVER removed from the Linux Kernel. You can choose to compile it as a module, or not compile it at all, but it's probably there to be compiled if you want it. Have you ever looked at all the possible kernel compile options? I swear you can still compile in options to support punch card readers if you look around long enough.

(Note: I'm joking a bit here, I don't know whether stuff is actually removed from the kernel source or whether you can compile in punch card support, thank you. But man there's a lot of option to support old stuff in there.)

Re:Or at least

[ManWithIceCream]

or just not use virtualization software made by a vendor.

Re:Or at least

[maitas]

Just use an Open Source virtualization Sofware (virtualbox, quemu ,etc. ) so whenever the vendor makes it obsolete you can recompile it by yourself.

Re:Or at least

[Just Some Guy]

Alternatively, go the matryoshka way. Run Win95 in (for Example) VMWare 5 on current Ubuntu now, wrap that in Xen on Ubuntu 12.4 LTS, wrap that in the 2016 Edition of Virtualooz on vanilla Lunix 28.6.19 and that in some deep fried beer batter. Processor speed will keep up.

IBM called that "z/OS".

Re:Or at least

[mr_mischief]

Those old applications under six layers of emulators are typically assumed to Just Work, too, and that's one reason why IBM mainframes are still sold for large sums of money. The I/O bandwidth compared to smaller iron is another.

Your father might be in for a shock

[try_anything]

Your father might be in for a shock if he thinks he can keep running the same computer system for the next fifteen years. Almost all veterinary clinics have a web presence these days (if only contact info, a map, and some cute photos) so it's a cinch that in five years the bar will be raised to include real online functionality. Make an appointment, see when your dog is due for shots, see how much Poo-Poo weighed at his last checkup -- sounds nice, right? His current customers won't care if he falls behind, but without a steady stream of new customers, his practice will dwindle.

That means he needs to plan on new software. Software upgrades are much more painful and expensive than hardware upgrades, and new small business software has a way of running poorly on five-year-old machines. The next fifteen years will bring painful changes for his clinic's computer systems, much worse than simple hardware upgrades, and he is the one who will have to understand and deal with it. Of course, he might soon have the option of having his data and applications hosted elsewhere, so he might be able to keep the same hardware for the next fifteen years after all, but I don't think that scenario satisfies his current expectations.

Re:Your father might be in for a shock

[Scrameustache]

a web presence these days (if only contact info, a map, and some cute photos) so it's a cinch that in five years the bar will be raised to include real online functionality. Make an appointment, see when your dog is due for shots, see how much Poo-Poo weighed at his last checkup -- sounds nice, right? His current customers won't care if he falls behind, but without a steady stream of new customers, his practice will dwindle.

I don't think there's a reason why a low-power system couldn't be used for data-entry to feed a remotely-served web presence.

In fact it would probably make more sense for a vet to hire an outside company to do his web services for him rather than do it in-house.

Re:Your father might be in for a shock

[try_anything]

I'm sure he will be pleased to learn that by replacing all of his software, migrating all his client data from a legacy system to a radically different web application, and retraining all his staff, he can avoid spending a couple hundred bucks on new hardware.

No, this guy would not be thinking about such small change if he knew what was coming. Unless somebody develops specific migration support from the old DOS (!) software he's using to a modern application with web support, software upgrades and migration will cost him tens of thousands of dollars of his own time and his employees' time, plus thousands in software and support.

Re:Your father might be in for a shock

[Gunnut1124]

Ok, for starters, both my wife's parents are Vets, as is my Bosses wife. I've been involved in quite a few talks with veterinarians about their plans for web-presence and their use of technology in the workplace.

You think modern hospitals are behind the times lacking electronic medical records and requiring warehouses full of paperwork, you'd be amazed at the antiquated systems used in vet offices. A 15 year old PC is not quite what my in-law's office uses, but they ARE still running Win98 on 10+ year old hardware. They needed a backup solution a few years back that demanded a CD burner... It was a nightmare. The installation process had a few moments where I was certain I'd have to tell my father-in-law that I had lost all his scheduled appointments and customer contact info. That all makes the point that, vets don't generally respect the power of technology, nor do they understand it's potential.

The entire lack of modern hosted services for veterinarians and just daily-use software is pathetic as well, but what can you do... they don't want to pay for the services, so they suffer the lack of options...

Re:Your father might be in for a shock

[try_anything]

You think modern hospitals are behind the times lacking electronic medical records and requiring warehouses full of paperwork, you'd be amazed at the antiquated systems used in vet offices.

Veterinary software will be way ahead of medical software when it comes to online access, because vets don't have to worry about medical privacy regulations. A web presence for a veterinary practice just has to be a nice secure site. Soon enough a company that develops veterinary billing and management software will hire a summer intern to add a simple customer-facing web interface, and from there it will be simple to add interfaces for more and more functionality until the entire thing is web-based and can be hosted off-site. Frankly, if I was developing veterinary office software, I would be doing that right now, just so I could offer a cheaper and easier solution. (No need to do backups, no stress or hassle when PCs fail, cheaper training for new hires who have never seen any program other than a web browser, etc.) Sounds like a perfect product for cheapskates.

Re:Your father might be in for a shock

[Gunnut1124]

what's that old saying? If you build it, they will come...

You find that software company and I'll have them 2 clinics lined up and waiting.

This seems like a perfect place to test medical software without the risk of HIPPA and the mess that follows... If i were a software dev, I'd likely have worked on this problem by now.... maybe one of the slashdot community already has?

Re:Your father might be in for a shock

[pbhj]

Are you proposing that he hosts his website on his own machine?

What software do you think he'll need other than a browser (if he decides to manage his customers online)?

Re:Your father might be in for a shock

[Anonymous Coward]

The real money in the vet business isn't Fluffy or Fido (unless it's a boutique vet storefront). The real money is Farmer Brown needing a couple dairy cows checked out, or other farm work. Rural area practictioners are much less in need of a web presence than urban ones.

Re:Your father might be in for a shock

[wvmarle]

I don't know what you expect but you don't need too much hardware for those two or three visitors a day. Googlebot will probably visit more often than all human visitors together on such a site. Just be realistic, if he now is using a decent database like MySQL at the backend then adding a simple web with user login and check-my-pet-status is really easy. He may be able to run it from a remote hosting company (but then the database updates may be an issue) or get a half-decent Internet connection.

This kind of stuff usually doesn't need much maintenance. As soon as it works don't touch the set-up and you should be fine for long time to come. I have to maintain my own web site so infrequent (and this site is database-driven, with my latest offers posted and removed automatically all the time, really more than some static pages) that by the time I want to change something I forgot how I set it up... it just works. You just have to be realistic and remember this is a small operation, really small, the slowest and cheapest modern hardware being a hundred times what is needed, this is not a website that gets hundreds of hits a second, ever. This is something many IT types tend to forget.

Re:Your father might be in for a shock

[wvmarle]

If it ain't broke, don't fix it.

Their job is to take care of pets, not to run a fancy website. The computer is a tool, nothing more than that.

Re:Your father might be in for a shock

[julesh]

it's a cinch that in five years the bar will be raised to include real online functionality. Make an appointment, see when your dog is due for shots, see how much Poo-Poo weighed at his last checkup -- sounds nice, right? [...] That means he needs to plan on new software

Speaking as a developer of such software, I'll say he can do it with his current system without upgrading. I'd make a browser-based booking system that works in IE3 and he can access over dialup. He'll have to reserve some slots to web bookings due to the fact that it isn't live-integrated with his existing system, but such systems are workable for most small businesses. Occasional unbooked slots can be useful to catch up with paperwork, and responding to a sudden increase in demand is easy enough as long as he checks the system at least twice a day.

Call IBM

[DoofusOfDeath]

I'm pretty sure you can get a Z-series mainframe that they'll support for > 15 years.

Re:Call IBM

[flnca]

The point with PCs is, that higher chipset and mainboard integration has led to less reliable hardware. Old 8-bit computers like the VIC-20 still run today without any problems (after more than 25 years). Systems like the Amiga were already more prone to failure (I have an A-1000 that died after more than 20 years). Old PCs like the old AT bus (ISA) 386/486/Pentium machines do still work (if the CMOS backup battery can be replaced). High-speed Pentium 4 boxes often have a mainboard failure within five years. Of course this depends on how well the system is protected from the elements like heat, etc. But definitely nowadays you can't plug just anything together and expect it to work for 20 years.

IBM offers i-Series servers (OS/400) for business applications which should have a long life as well. p-Series (AIX) servers aren't bad either. In any case, IBM hardware should be more reliable than PC hardware. I'm sure they offer spare part storage as a service as well (if it's not already included in the price). But IBM has a reputation for abandoning its products. I don't know if a proper contract can protect a customer from that.

He likes what he has right?

[freedom_surfer]

Just build a new single system and virtualize the two components of the previous system. He won't have to learn how to use it, you know it works for what he needs and considering you are emulating such old hardware, can be done on cheap modern hardware. Just get a low end core 2 duo with VT, and install Ubuntu and kvm virtualization. You could probably get a bare bones system for the same price as two new Microsoft licenses. Spend the real money on a big display for the old man instead!

Ridiculous Replies

[Anonymous Coward]

So many of these replies are utterly ridiculous. He is asking for a replacement for two 66mhz workstations with 500mb of storage to run an old dos app and people are suggesting raid 5, xserves, mainframes, etc? This is a "veterinarian with a small private practice" -- I think he will survive without ten remote backup sites and eight firewalls. It doesn't even sound like the machines are internet connected, why would they be?

Give up the 15 year fantasy. He got lucky, he might again, but don't count on it. I'm guessing he wants a system that will last this long mainly to save money. Well, he's in luck, computers are a lot cheaper nowadays. He can replace it every few years and still spend a lot less than he originally did.

My suggestion is just to get some cheap desktops from any well known beige box vendor, install vmware/virtualbox/virtual pc, periodically back up to a thumb drive, and be done with it.

IINBDFI

[Anonymous Coward]

If its not broke...

Why change anything?

Win95 needs to be nuked and re-installed every few years. If this machine is not connected to the internet, why not just clean the detritus out of the hard drive, or start over with a fresh install after a proper backup of the hard drive, and maybe a new SMALL hard drive.

You will have more problems getting that old DOS system running on anything you can buy today.

You might upgrade to windows 98 or Windows 2000, but getting old dos applications to run on current versions of Windows and current processors is often a real pain.

Re:IINBDFI

[Gible]

What he said

Also, if cleaning out the crap on the hdd(of which there may be little), isn't enough, throw in a bigger, old hdd. from memory win95 can handle up to 2G. And like his system, if the old hdd still goes, then it'll probably go for a fair while yet. If 500MB lasted him 15years, then quadrupling that should give him another 45 years or so.

Re:15 years?

[blueg3]

What purpose is there in spending money unnecessarily?

Re:Simple answer, is it's time to redesign the sys

[Anonymous Coward]

Postgres sucks. You should look at MySQL, which has enterprise-level abilities. Don't trust your data to some small hobby project like postgres.

Re:Most 15-year old Sun workstations are still use

[downix]

I run an Ultra 10 and I concur with this statement.

Virtualization

[drinkypoo]

Just take image-based backups. When the system finally fails, virtualize a new one on top of whatever system can do it cheaply and reliably. Repeat until there's some compelling reason to upgrade the software, and thus the [virtual] hardware. Start over with a virtualizable platform... Right now your best bet is Linux on x86, but Windows on x86 is a strong contender as well. x86 should be around for quite a while yet and if it isn't then you can bet there will be shitloads of emulators.

Electrolytic capacitor problem

[prefec2]

The electrolytic capacitor on the main board are also a typical part to fail. The hotter the system the shorter there lifetime. So a cool motherboard and system is required.

Re:Electrolytic capacitor problem

[davetv]

I've lost a couple of motherboards due to electrolytic caps failing. I'd suggest a mobo with "solid state" caps.

Re:Electrolytic capacitor problem

[Anonymous Coward]

Or use a motherboard that uses solid state caps.

Re:Electrolytic capacitor problem

[julesh]

So a cool motherboard and system is required.

Yes, indeed. Underclocking is your friend: it reduces heat and power consumption, and extends the lifetime of most of the components on the board. Hard disk reliability will be the biggest issue after that, so a RAID1 pair with a spare in a box somewhere so you can just swap it in on demand seems the best choice.

Re:Electrolytic capacitor problem

[frozen_kangaroo]

Get the soldering iron and just replace them ! I've kept a couple of motherboards alive doing just this. They let you know when they have gone because they split open at the top.

Re:Electrolytic capacitor problem

[PitaBred]

There are actually a lot of motherboards sold any more that have solid-state capacitors instead of the electrolytics. Usually pitched to gamers, but any hard-service machine would do well with them.

JPL

[Tablizer]

Ask the Mars MER rover team.

Re:JPL

[Leebert]

Ask the Mars MER rover team.

Wow. "ATM Machine" just doesn't do it anymore. I now have to refer to an "Automated ATM Machine". :)

Pentium 66mhz and a 486, both with 8MB of RAM

[linhares]

Well, my friend, with those specs, you may have to consider that your father may fail first.

Mine did.

Re:Simple answer, is it's time to redesign the sys

[QuoteMstr]

Sometimes it's hard to tell the difference between fanboyism and astroturfing.

IBM Mainframe

[Anonymous Coward]

Get an AS/400 they will always be backwards compatible. In my experience they are the only machines that can go for 15+ years without any issues. A tape drive for backups is must too.

Consider changing IT cycles.

[barfy]

There are two possible solutions.

First is to change nothing. Why fix what isn't broke?

The second is to change your time frame entirely. 10-15 years is too long and too disruptive when the time comes, and you lose out on presumptive benefits in the middle.

Surely there are network aware applications that do what you want on standard systems today.

You want to be network aware. In todays world you do not want to be cut off from your customers, and more importantly you want to push of data integrity to others.

You should develop an annual budget for IT expenses that rolls over. You should be on a 3-5 year schedule rather than a 10-15 year schedule. If you do this, you will have more predictable costs. You won't have competitive disadvantage because of software. You will have advantages of providing more and more reliable services to your customers.

As in all businesses information and digital information can be used to extend and monetize your business in all sorts of ways. But only if you choose to keep on top of it, and you don't constrain your learning cycles to whatever is new now.

Re:Consider changing IT cycles.

[Anonymous Coward]

WTF are you talking about. Your entire response is a business school alternate reality filled with buzzwords that do not relate to the original requirements. It doesn't need to be network anything, he certainly doesn't "want to push data integrity to others" and he certainly wants new hardware.

The current system has been running for 15 years and has almost filled up a 500MB hard drive. He can replace the entire box with a [cheap, lower power] setup running on a RAID-1 of as many flash memory cards as the mobo will support.

Not everyone wants to be ready for Web 3.0, part of Web 2.0, or even on Web 1.0
Sometimes you just want a computer that does one thing well and doesn't interface with anything outside the room.

Build yourself

[linzeal]

Zeroth is of course fans, they are the only moving part so the only thing I use is dual ball bearing fans.

First thing is quality of motherboard I would go with Solid Caps, heatsinks (NO FANS) on the S and N bridges and look for an 8 layer PCB design they tend to be more robustly engineered.

Second is to underclock and undervolt the CPU and memory which will increase the lifetime. I have an Athlon 5000+ running at ~1 volt that runs at 24c in an almost fanless case. He is not going to be using it for anything that requires a lot of clockcycles so pare them down as much as the CPU/Memory will allow while being stable. It will increase the lifetime dramatically.

Third you need an all metal computer case for durability. Everything that is plastic in 10-15 years will become brittle and stuff will start breaking like around the power switch and other heavily used areas.

     

No way...

[Anonymous Coward]

Wow... your advice might be somewhat interesting in the general semse, but I think it totally misses the point for this particular question.

The guy has a legacy DOS app he wants to keep running. It runs fine on a 486 computer. Redesigning this system to be a quadcore server with new webapps etc is crazy! (not to mention, web technologies are about the LEAST stable area of computer development at the moment). The guy is running a 486 with a 500mb harddisk, and you want him to upgrade to a quadcore server with tons of memory and room for multiple disk expansions? I just don't agree with this at all..

I also doubt that in 15 years SATA drives will no longer be available. Hell, I doubt that IDE drives will be that hard to find in 15 years, though I could be totally wrong on that one...

The solution--IMHO of course--is what many others have suggested. Virtual machines.If you need two computers (as the poster mentioned), and REALLY do not want to worry about any changes in setup for 10-15 years, then buy three or or four modest desktops (and I would highly, highly recommend a backup server that does regular backups and lives offsite...this can run linux or whatever floats your boat). Design the system so that if one computer dies you can swap in another painlessly (restore from backup, unplug old, plugin new, done). Beyond that, UPS UPS UPS. IMHO, even with reliable component brands, getting a computer to last 15 years is still a crapshoot. I've done it with crap generic brands, while good brands have had unexpected failures. Plan for eventual hardware failures and you won't be disappointed :)

I've been in a similar situation maintaining servers for a small family business. Depending on your age now, you may have a lot of time and flexibility to help the parents, but if you end up moving else where, college, gradschool, new job, whatever, you want the system to be simple and to take as little of your time as possible! It's best for all involved. (also consider VNC/Remote Desktop/SSH/etc to allow you to help out remotely)

Virtualization and Lapkosoft

[Anonymous Coward]

For less than 1 grand,you could get several used,refurbished IBM servers that would easily last 10 years and have all the spares you'll ever need for the other 5 years. Set it up with a recent copy of fedora and you'll have all your father will ever need for his vet service

A great solution!

[beaststwo]

I'm working with a customer that has a very high regulatory sensitivity to hardware changes. Whenever they change the hardware platform, there are mountains of paperwork to fill out.

We went with Vmware,so that as they have to change hardware dueto life-cycle replacement, the virtualized hardware seen by the OS and applications never changes.

Either Vmware's or Microsoft's products should provide as consistent a virtualized environment as can be had today. Certainly less painful than moving the existing software environment to a new machine.

System i

[PoiBoy]

Move everything to an entry-level IBM System i machine with a couple of terminals. The hardware will last 15 years and the OS is as reliable as a tank.

http://www-03.ibm.com/systems/i/?cm_re=masthead-_-products-_-sys-iseries

don't sweat it

[the_wesman]

it sounds like me like those are run-of-the-mill PCs from 10-15 years ago that still meet his, limited, needs for work. he's not gaming or interwebbing or anything. Just buy him a run-of-the-mill PC from today and he'll still be using it in 15 years. 15 years ago everyone was freaking out that computers are "going to go obsolete as soon as" you buy one - right. it's not really that - it's that needs, and program's requirements change. I've got a friend that has one of those giant, beige desktop computers (before towers were all the rage) that I laugh at when she mentions it, but she's able to check her email on it and she's got a computer at work for doing work. Other than that, she has a social life, so the computer is fine for her. Obsolescence, with computers, is likely more about boredom and shiny new things than it is about technical limitations.

No Moving Parts

[AvitarX]

I would pay for all solid state devices and no moving parts (or a case with one or two large fans).

I would get a solid state motherboard, a few solid state HDs, if I were using Linux I would software RAID 10 them as many copies of data as drives, this should allow for multiple failures, and no need to replace. I imagine Windows has something similar.

I would get a CPU that can cool without a fan (under-clock with heat sink, heat pipe if needed).

Get a solid state power supply that will be nowhere near it's peak usage.

This will probably double, triple, or even quadrupole your cost, but if you want 1994 system reliability, plan on spending the same amount a system in 1994 cost.

Just Buy Quality

[digitalgimpus]

Quality counts. IBM is known for long lasting systems, and having a warehouse with every part for every damn thing they ever made. Lenovo's still got enough integration and legacy from IBM that this still seems to be the case. I've also got old Mac's. My last home server was a Beige G3, my current one is a B&W G3. I upgraded since I had the spare system and it had room for more drives. The beige was still running. That B&W G3 is now 10 years old. Parts most likely to fail are HD and fans/psu. All are easy to replace with generic equivalents even on old systems. Not even that much $$. Bottom line is buy quality. The parts that will most likely fail are the easy ones. It's the motherboard and CPU you care most about as they are most likely to become sparse.

Backward compatibility

[gmuslera]

The problem is that needs evolve. The core could still be that DOS application, but what is around it could be taken advantage too (even if is just taking advantage of new widely available hardware or internet)

It could still be running his application under DOS, as long as there is nothing very specific/hardware dependant on it. Probably could run unchanged in vmware/virtualbox, dosemu/linux, newer windows, etc, and even backed up in a pen drive for portability/backups. The application will run in whatever machine the current other needs will require.

P2-333/300 still chugging along after 11 years...

[logicassasin]

I have under my desk an IBM PC 300GL desktop and an IBM Thinkpad 570 which I take to work with me every day. The 300GL is a P2-333, and 570 is a P2-300. Both are in use daily (especially the laptop). The 300GL has 256MB of RAM (the max the board will accept) and the 570 a paltry 192MB (it's max). Both run XP SP3 only because I need it for my cameras (no win2000 drivers). Both machines are good for basic stuff: Email, word processing/spreadsheets (via Open Office), web browsing (using Opera), and can handle photo manipulation if a bit slowly (The Gimp). With a bit of help, I can even watch some Xvid/divx encoded movies (MPlayer for Windows) but I have to resize the video in the files using Avidemux on my main PC (Athlon XP 2500+).

For a large percentage of people out there, these would be fine for daily use. I would imagine that in 10 years, a PC bought today would still be quite useable.

Buy Apple

[gjyoung]

Buy an Apple. my wifes iMac lasted ~10 years before we replaced it, and the only reason we replaced it was for a bigger screen for her. A current top end Mac Mini and a 20" Dell LCD will be relevant that long easily.

Re:Buy Apple

[Ash-Fox]

Buy an Apple. my wifes iMac lasted ~10 years before we replaced it, and the only reason we replaced it was for a bigger screen for her. A current top end Mac Mini and a 20" Dell LCD will be relevant that long easily.

Providing more anecdotal data - Over the years I have had at least, at least twenty Macs, including a MBP two years ago.

Every single one had at least one logicboard failure.

Convert to a virtual machine

[jschimpf]

(1) Convert the machine(s) to virtual images.

(2) Host on Linux boxes/generic hardware

(3) Arrange to boot into the app running in the VM

(4) Backup the images regularly

(5) Buy a spare box

The VM image is effectively immortal if hardware fails just buy whatever is cheap & current install Linux and the VM program.

My 2 cents worth...

[jmccarty]

For what you are trying to do, I would suggest just spending $1-2k on a nice new middle of the line workstation. Get an LCD monitor, Laser printer, and just modernize all of the hardware. Migrate your customer database to something newer, such as QuickBooks,and buy a few flash drives for regular backups. That way your data is protected, and you won't need to spend a bundle on fancy hardware. If something breaks, you can bring it to a PC repair shop, and your data is not stuck in an out of date format.

Re:Most 15-year old Sun workstations are still use

[blind biker]

If you have Solaris 10 on that Ultra 10, how much RAM does it require? I wonder if I could get away with 512 MB.

Is the harddrive *really* filling up?

[Rix]

With your father's data, that is, rather than Windows 95 just barfing over the drive.

I pray this system isn't connected to the internet in any way, because if it is it must have hundreds of worms crawling around in it. Windows 95 is of course terrible for this, but any system you plan to keep running unmanaged for 15 years should be kept far from any network and physically secure.

I really can't imagine a single veterinary practise generating gigabytes of administravia, nor can I imagine some slapdash DOS application that does generate gigabytes of superfluous data being able to index it once it grew to that range.

Check to see how much he's really using. If it's a small enough, move it over to a flash disk and run the application with DOSemu or in a VM. Build a system with cleanable/replacable air filters over the fans, and train your father to back up his data. (If he hasn't had a hard drive fail in 15 years of use, he's damn lucky.)

Re:Is the harddrive *really* filling up?

[gatkinso]

Would a contemporary worm actually work on Windows 95? I can't see a worn author supporting anything earlier than NT 4. But maybe they have better tech support than most software companies. ;) (Actually they probably do!)

Security thru obscurity/obselescence.

Re:Is the harddrive *really* filling up?

[Just Some Guy]

I pray this system isn't connected to the internet in any way, because if it is it must have hundreds of worms crawling around in it.

He's a vet - they have stuff to treat that.

Re:Is the harddrive *really* filling up?

[ndazza]

As the OP mentioned, the hard drives are 500 Megabytes, not gigabytes. A few hundred word documents with the company logo in it will fill that up in no time.

If you love your father...

[Fortunato_NC]

...you will tell him to get a professional to do this for him. Whether he understands it or not, your father's livelihood depends on having computers that are up and running, and from the limited picture you've given us, it sounds like he only thinks about information technology when something is going wrong.

Simply replacing his current hardware with newer gear is just kicking the problem down the road. In the last 15 years, there have been significant advances that he should consider taking advantage of, because they can make his business run more efficiently. But the only way to determine the "right" solution is to have someone who understands your father's business design a solution for him. A good place to start would be with the vendor who sold him his current setup, if they are still in business. They can most likely recommend a suitable hardware platform and assist with data migration to a newer, supported version of their software, and provide some sort of service arrangement that will ensure that these systems are maintained, not just used.

If you're bound and determined to do this yourself, the recommendations about virtualization are good ones - you can build a couple of VM images that you can backup to a USB flash drive nightly, or even better, several flash drives - learn about backup rotation schemes and design one that gives you the ability to recover two weeks worth of image data at least.

Up thread, someone said that if you do this, you will be supporting it for the next 10-15 years. You need to keep that in mind when deciding how to proceed. Best of luck!

Re:If you love your father...

[Dahamma]

YES! Agree with your post wholeheartedly.

I feel like I wasted my time with an elaborate post below explaining my own father's veterinary practice that will never be read, when I should have picked a better title like yours ;)

Re:If you love your father...

[Ash-Fox]

you will tell him to get a professional to do this for him.

You know, in my experience there a lot of crap professionals, despite the fact I am quite good in the IT field - even I have problems finding good professionals.

Re:If you love your father...

[Anonymous Coward]

be very careful "hiring a professional" and get/check references. computer consulting isn't a regulated profession, and I've met some consultants I would trust to wash a car without breaking something.

the big variable is software. if the current app runs ok after Y2K, will it still run in 2039? 2050? Date-time arithmetic is the killer for legacy software. If he can't schedule new appointments because the software thinks next year is 1980, that's a problem.

IBM still sells DOS. Call them on the phone, tell them your troubles and ask about supported systems for PC DOS. IBM doesn't make PCs anymore, but Lenovo (who made Thinkpads) still does. Basically, you slow down the CPU in the BIOS (if necessary), install DOS, and go. The application may have trouble with large hard drives too. It may report 0% free space on a 40 gig hard drive, unless you keep the C: partition under 2 gigs.

Or you can help him shop for new vet/medical office software. Then help him get the hardware. Shop around; it may not cost so much and he will have a vendor to call (instead of you) if he has requests.

Or you can write him a custom system in FileMaker Pro or Access. Don't do this.

Re:15 years?

[LWATCDR]

Not cheap but smart.
You say cheap but he didn't ask for a cheap system that would last for 15 years.
Replacing systems every few years is a pain. It costs time and money. Not only that but you are filling up land fills with machines.

I vote...

[prometx42]

...virtualize. I wasn't even thinking that until I read jonbryce's comment, but I think he is correct. Build yourself a nice rock solid, inexpensive, cool running multi-core system, say ASUS MoBo, and AMD Phenom CPU. Slap a nice, rock solid Debian or Ubuntu no-frills install on there, and virtualize the database and it's software front-end. Though I would bring him into the XP era with the virtual machine. Put in a nice DVD-ROM drive, so that he can backup to disc. Any current series mobo will have usb out the wazoo for further backup options which will be backward compatible with the upcoming USB 3.0 Standard, which should cover forthcoming peripheral devices for backup going forward. Noting that this machine won't have to run the software of the future, but is actually task specific to running the software of the past. I would not stockpile hardware, just keep the data rigorously backed up and play the hardware market going forward. Good luck with that, sounds fun actually. Cheers...

Re:I vote...

[Danzigism]

I couldnt agree with you more.. I've done some work for hotels in my area that used to run some really old applications up until about a year ago.. There are still records of information that needs to be accessed on that old system from time to time being that they had to start off fresh with their newer more modern booking software.. thanks to virtualization, they can access their old info with ease since their entire database and their program can literally fit on a CD and booted up using VMware Player.. I think for this particular purpose, I would almost considering getting newer more up-to-date software that can give you and your customers more.. even though the DOS stuff works, the data is what is most important.. it seems that the larger the database gets on those old systems, the more chance they have of becoming corrupt.. preserve it while you still can, keep it around for your records, and considering a full upgrade of both software and hardware with a good backup solution..

LISP

[janwedekind]

Port the application to LISP. In 15 years everyone will see that it was the programming language of the future all along.

depreciated cost

[O_Sleep]

Try to sell your dad on a spending rate. You can't guarantee a computer to last 15 years, but if you can figure out his risk tolerance and try to maintain his spending allowance. If he bought two 486s in 1994 and each were $2000 and the value of the dollar being about 1.3 what it is now, so he spent $5200 in today's currency back in 1994, that's about $350/year. Do you think with these numbers, you could convince him that it would be worth buying two netbooks for about $350 each, with the high chance they would last two years? If they last an unheard of 3 years, he would actually be spending less per year 700/3 ~ 250. If they last more he will actually be saving money over his old setup. In fact, if he took the same amount of money, $5200. Invested $700 in hardware. Put the remaining 4500 into bonds at 5%, over 15 years he could almost buy a new netbook every year on interest alone. -Bjorn

hmm 2nd stupid question of the minute

[nibbles2004]

2 question's, veterinarian's are not poor by any standard's so to invest every 15 years in crucial infrastructure is just being cheap, and secondaly how old is your father, will he actually be working in 15 years, i suspect he be in bognor regis by then, so don't worry about it.love

Re:hmm 2nd stupid question of the minute

[try_anything]

how old is your father, will he actually be working in 15 years, i suspect he be in bognor regis by then

His father owns a share of the practice, or maybe the entire practice, and will sell his share to another veterinarian when he retires. Even if he has no sentimental attachment, he has an interesting in ensuring the future of the practice so he can get a good price for his share.

Re:hmm 2nd stupid question of the minute

[try_anything]

he has an interesting in

Sheesh, time for me to quit for the day.

Re:hmm 2nd stupid question of the minute

[ErikZ]

Interesting.

So you think by a 10-15 year old PC would be an asset instead of a bigass liability?

virtualize

[Dale512]

The person above that mentioned virtualization probably has the best idea. We have a machine at our office that is 8 years old. It is not connected to the internet or network. Its sole function is to run our dated door lock system. The software runs on Windows 98. It will not run on an XP machine. The upgrade to the software is $3,400. Management is not willing to spend that money just to lock/unlock the doors. We have no backup and it took us bit to get a usb drive up and running on it so we could back anything up. Ultimately we decided to get a new xp machine for it and just run a virtual instance of Win98. It works like a champ and we can simply save the virtual state for backup to a file.

Re:15 years?

[poopdeville]

Driving the Western, credit-based economy.

a couple of options + prices

[CAIMLAS]

There are a number of options; I would say that your least favorable option is remaining with dated hardware. Frankly, it's nothing short of a miracle that 2 15 year old systems have lasted as long as they have without failures.

1) remain using dated hardware. Not the best idea; eventually (possibly sooner than later, with such old rotational storage) you'll run into a situation where old, unused drives will die quickly due to the drying of the lubricants.
2) See if you can't figure out a way to run the same hardware with CF -> IDE adapters. In this case, you'd have a marginal cost increase. You'd still want to keep 1-2 systems' worth of old junk to replace things as they fail.
3) Virtualize on newer hardware. This is, IMO, the best option.

It would seem to me that the primary requirements are:
a) prolonged system longevity
b) static software

In these two situations, you'll have a couple of options, which I will address; but first let me mention similar experiences I've had.

I've dealt with this issue a number of times with my father, who runs a small business out of the home. Every 2-5 years he will get frustrated with how slow his computer is getting (in his case, because he installs a lot of crap) or his hardware starts to fail. Invariably, he will insist that his computing platform should remain static - until I forced him out of it, he was still using a (very crappy) program called Act! 2000 (and he'd still be using W98 and Office 97 if I hadn't forced him out of that years ago).

The fact is, newer hardware simply isn't supported by the older, abandoned Microsoft operating systems. There's no two ways about it. You can't keep using old hardware, even if your software is static, if you want to be certain the information will be available in the future.

If your father is like mine, a big part of the "make it keep working" reasoning is that he doesn't want to have to buy anything new. The reality is that you could replace his systems for a fraction of the cost of what he originally paid for one of them. It's a business, and $500 or so in cost every 15 years is NOT an unreasonable proposition. Cheap is one thing; technophobic to the point of disaster is another.

In my father's case, there was a very sound reason for upgrading his software: it was incompatible with newer file formats. If you've got a closed system (like a POS or accounting package that is the only software you use, as appears to be your case) there's not much of a need.

If you're looking to have systems run for a prolonged period of time, you will want to eliminate potential points of failure as much as possible.

This is what I would do:

* Get two new Via Nano (or similar, such as this MSI Wind barebones for $150) systems with cheapest-possible LCD monitors. IE, on the very cheapest end of things, but newer and all-around better than what he's got.
* Do not use rotational media for these systems. Get a 4Gb or so CF card with an SATA or IDE adapter for each. (yes, the CF will hold up to use just fine)
* Stick 1G (or whatever's cheapest, really) of RAM in each. Get a couple extra sticks, just in case, and keep 'em on hand (because if htey need replacement, it'll either be immediately after purchase, or right around the time when DDR2 gets pricey due to lack of demand/manufacturing).
* Buy an $80 UPS with line conditioning. This is necessary to encourage the hardware to not fail due to a spurious power brownouts and the like.
* Buy an extra USB flash drive of twice the capacity of one of the systems to back up the systems and data - either automagically, initially, once a month, etc. - whatever is appropriate for his use patterns. At the very least, back up your initial install.
* Get VMWare Server (either v1 or v2 - I'd say v1 would be preferable due to the console it has) and install your W95 machines on them.
* Set up each machine to automatically boot, log into X, and launch the VMWare Server c

Re:a couple of options + prices

[Ash-Fox]

There are a number of options; I would say that your least favorable option is remaining with dated hardware. Frankly, it's nothing short of a miracle that 2 15 year old systems have lasted as long as they have without failures.

You know, I still have a bunch of Amigas, some them going back as far as 1994, none of them have failed yet, in any form and I do use them regularly. That said, they're fanless systems. I am surprised the hard drives haven't gone yet, actually. I should probably look into replacing them with SSD.

Re:a couple of options + prices

[Ash-Fox]

You know, I still have a bunch of Amigas, some them going back as far as 1994

Sorry, got confused, I meant 1987, not 1994.

keep the existing setup and virtualize it

[Bill Wong]

keep the existing setup and virtualize it (vmware is nice, but, xen is free)
upgrade the vm host hardware as needed or as necessary, instead of upgrading a machine that should really be left alone.

benefits:
1) minimal effort needed on your part
2) your dad doesn't need to learn new software as it is exactly the same, and all the data is retained
3) hardware upgrades are transparent to the client vm (hopefully)

(but, don't forget to routinely backup the data on the vm like any other machine though)

Re:keep the existing setup and virtualize it

[Pentium100]

Last time I checked VMware Server was free too.

Re:keep the existing setup and virtualize it

[Bill Wong]

Free as in beer, I guess. :3
You're right though, i completely forgot about vmware server being free.

Re:Most 15-year old Sun workstations are still use

[Anonymous Coward]

Forget it. You won't get enough spare parts or they are too expensive.

Lost two Ultra-1 due to hd and fan failures -- both 12y of age.

Convert to FOSS

[fast turtle]

First thing I'd look at is moving him over to a FOSS solution such as Linux/*BSD. You also need to look at converting his database from its current format over to either MySQL or Postrgress for longevity because when the software is no longer supported, things will become a nightmare. The next thing I'd look at is using KDE 3.5 with the Redmond settings (works like Windows) even though the icons are a bit different. As to his Accounting app, look into either KmyMoney2 or GnuCash as they both use an open standard format (Check Quicken/QBooks format import of both).

Re:Most 15-year old Sun workstations are still use

[aphaenogaster]

have you looked at the cost of a new sun machine? Very inexpensive. Anyway, if sparc were an option I would get a blade2000 off of ebay. my blade1000 is 11 years old and has never had an issue.

My list, FWIW

[Anonymous Coward]

*Hard drive. It was a high-quality drive, but it failed after 7 years. Good hard drives may last longer, but there is a strong stochastic component to drive longevity. You can't guarantee 15 years. No way.
*Fan on video card. I didn't replace it because the video card didn't get hot, not even when playing 3D intensive games.
*Mouse. 5 years, I bought it later. Hand-on-mouse detection sensor failed because of shoddy construction, so you may get a better life time, but íf you're wise you'll prioritise ergonomics instead.
*CD burner. 8 years, software failure. Generic drivers didn't work with the burner, and the official drivers failed after an OS update. The manufacturer is belly up, so I bought a new drive.
*Keyboard. 8 years, but I wasn't very nice to it. On inspection, the cause was probably loss of conductivity in a signal line due to corrosion. But I discovered the design of the keyboard was particularly vulnerable to this, so a good keyboard will last longer. Or just be nice to it in the first place. I also have working keyboard from the 80s but it isn't very comfortable and it makes enough noise to drive you around the bend and down the sewer.
I think you should tell him that a) his system only lasted so long because he got lucky and b) that since you can't guarantee such a long lifetime, you won't design for it because you would be making a promise you can't keep.

The safest solution

[The Famous Druid]

With a hardware solution, the nightmare scenario is that some crucial component fails, and the business is off-line for days/weeks while you try to source a replacement for a 10 year old part.

With a virtualized solution, if the hardware fails you can buy any new PC at the local Wal-Mart, restore your backup of the VM, and the business will be up and running in a matter of hours.

TRWTF

[timmarhy]

TRWTF here is he is doing work for family, even worse family which clearly has no idea.

15 years for any equipment is exceptional, not the norm. any spastic knows this. either backup or virtualise the current system if he is happy with it, and install on a $500 dell or something similarly cheap.

remember hardware is getting cheaper all the time, so your better off going cheap now and buying a $500 one every 5 years than buy a $1500 one thinking it'll last longer.

Not totally infeasible, actually.

[Junior J. Junior III]

The two components most likely to fail on any system are the hard drive and the fans. Next most likely to fail are the power supply and the capacitors on the motherboard, graphics card, etc.

You can build a PC using a SSD, and then you won't have to worry about the HDD failing. SSDs have a ridiculous theoretical MTBF. If they actually live up to the figures I've read, 15 years shouldn't be a problem.

Power supplies are cheap, mostly interchangable, commodity parts, easily replaced if a failure occurs. Some PSU failures will take down other components with them, though. If you get a good brand and don't put too high a demand on it, it'll likely last quite a while.

As for the capacitors, if you shop around you can find motherboards that use solid capacitors, which are supposed to last longer.

Fans are relatively cheap and easy to replace. You can improve reliability by going with low-power parts that can get by just with passive cooling. Chances are, the only two fans you'll have in your system are in the power supply and on the CPU. You might even be able to go with a passive cooling solution for the CPU if you have a beefy enough heat sink and use a low-power chip like Atom, and/or underclock.

They're starting to make LCD screens that have LED backlights. If you buy one of those, it should last quite a bit longer as LEDs are far more reliable than CCFL backlights.

None of this will guarantee that the system will last 15 years, but it'll give you the best chance you've got.

Another real problem is staying off of the upgrade treadmill. If you keep the system off of the internet, and never need to upgrade software, that should help. If you must be on the internet, security patches and other software updates will eventually bloat your system up over time to the point that it won't run like it did when it was new, just because the hardware demands of the new software versions is so much higher.

A lot of people regard running on the same hardware for 15 years to be silly, but advances have leveled off a bit, and it's feasible to run a system like Windows XP, which is already 8 years old, on the same hardware that you build today, for a good, long time if you don't need it to be on the network.

Speaking of depreciation

[Jane Q. Public]

The IRS will let you depreciate your computer hardware over a period of no less than 5 years. Regardless of the fact that 5 years is unrealistic for a lot of computer hardware these days, I would tell the man to depreciate them properly for those 5 years, and not plan for them to last longer than that.

Todays new tech is unproven.

[w0mprat]

Todays SSDs are completely unproven at 10-15 year life spans. They may not significantly outlive a good mechanical HDDs, as there is no proof for another 5-10 years.

Sure, there's plenty of hardware around that is 10-15+ years old that runs nicely. However I have a uneasy feeling in my gut that most of today's consumer and enterprise grade hardware is not as heavily built as stuff once was.

My concern being it is known to be more cost effective in certain situations to replace a certain percentage of hardware under warranty, than it is to build all hardware to a more durable spec out of the factory door. Oh yes and planned obsolescence too.

heyhey

[Anonymous Coward]

wow. are you serious? im working with windows vista which is several generations ahead. a new laptop cost 700-1000 dollars is very good already and with a few software you get your dad up to the 21st century. i feel like im talking to an someone in the old days. lol. just kidding.

I doubt that it's about price...

[jopsen]

It's more likely that fact that he doesn't want to learn a new system every 5 years... He just wants the things to work... And probably couldn't care less for fancy graphics and features.

Re:I doubt that it's about price...

[nedlohs]

You don't have to upgrade the damn software in order to replace failed hardware.

There isn't a hard drive on the planet that is warrantied let alone guaranteed to last 15 years. You can get lucky, but you can't build a system that will last 15 years with off the shelf (and therefore affordable) components.

I guess if you had to try you'd find the slowest RPM drive you could and cross your fingers.

A more standard system running whatever virtualisation system you prefer to run the ancient software seems the most practical solution, should be able to drop in a replacement machine/part if something goes wrong, and makes backup very simple (it's just a image file/directory/whatever).

Re:I doubt that it's about price...

[jopsen]

You don't have to upgrade the damn software in order to replace failed hardware.

That's right... Now we just need to find a highlevel programming language and framework that won't be obscure in 15 years... As far as I can see that's not easy, and C is not a highlevel language :)

Build it yourself

[Ponder Stibions]

Easiest way to get good quality parts where it matters. Like others say the HDD will fail first, so make sure you get a server grade one, not a consumer grade. Otherwise, stick to standards, go to the new RAM you can afford, and that can be stockpiled as it will fail and in 5 years they won't be making it, so it'll cost a fortune. Keep to the most standard parts you can, then when they go you have the biggest selection of replacements. Get quadcore compatiable boards so you can run dual or quad etc, keep your options open when you need spares. Software wise, as so many have said run Linux based, ideally start on a stable distro and a long term support release too. Something really old like Debian, as even in 10 years time it will still have followers. Even if the support does disappear, if it's open source you can fix it yourself or find some who can (even if you have to pay!). I would also suggest getting the software to run via another layer, so it could run on almost anything. A web based app would be ideal, but even a Java app would be better. Then as long as you can run Java and the machine is fast enough, you can run it! No OS worries. Run a server and have netbook clients, as mostly they won't need much processing power! When they fail, just replace, hopefully you use something that gets over the OS issue.

Replacement parts?

[gollito]

One thing to keep in mind with these type of machines is replacement parts. Back in the day those parts seemed to last a lot longer. Now days it just feels like they wear out quicker. Make sure you can find replacement parts (72 pin SIMM anybody?).

Re:Replacement parts?

[gatkinso]

I have about 2 gigs in 72 pin SIMM memory of various variety. Some 30 pin too. what do you need that you can't find on ebay? Hell I'll send it to you for free assuming US or Canada. It is taking up space on my shelf, and throwing it away would be a waste.

NVidia Tegra

[im_thatoneguy]

People are going the wrong direction. Don't build 1 big beefy machine.

Build an incredibly small lightweight machine. Like the NVidia Tegra platform. It costs less than $200. At that price you could replace it every other year for 20 years for the cost of one low-end server.

Now. Find a simple web-app which can be hosted. HTML and CSS don't change from decade to decade. Use a CSV file as the database. Now setup a one click option which downloads the data to a CF Card.

This website should be entirely file based so no need to install php, cgi, mysql.. nothing. This way you can copy an entire clone of the site to a CF Card.

Make sure it's a password protected ZIP file that is downloaded so that customer information can't be stolen.

Install a fast booting linux distro with Apache and Google Chrome or Firefox.

Now all you need to do for 15 years is install the latest web browser. And backup the website to a removeable drive every couple of days.

Here are some ways to add bonus points: Put the OS on a CF Card and clone it. Since all he needs is a dumb client and a barebones web server you should be able to fit it pretty easily onto a 2GB CF Card. You also now don't need any fancy ghostingesque software just a straight drive image. Keep 2 copies of the OS on CF Cards. Then if the computer ever breaks. Just replace the CF Card. Have a button which copies the latest website state to the new OS and off you go.

Websites are designed to be portable. They're meant to run on new hardware without a lot of setup. They're the perfect candidate for an application which can move easily from system to system and since you're only spending $200 on hardware which has no fans and no moving parts, but insane redundancy and replaceability I think your Dad should be able to run it himself.

"If it stops working. Just swap out the 'broken' card with the 'new' card." Then every 5 years you can update the instructions on cloning his system drive using the latest software.

And you've already written a little one click macro to restore and backup his web data so he should be able to easily follow the step by step instructions.

I could see this system easily running for 20 years.

Re:NVidia Tegra

[im_thatoneguy]

Oh yeah one other thing. That CF Card that he backs up his data to should go home with him so that it's an offsite backup.

Also it doesn't have to be a CF Card. Obviously now you could very easily use a USB thumbdrive or an SD Card as well. But since all it does is just copy the data from a portable drive to the apache www folder the little macro could easily be updated to meet whatever new portable storage medium is king.

Re:NVidia Tegra

[Anonymous Coward]

I guess it is okay for a vet to take client and patient data home on an SD/CD card ...

Personally a red button on the desktop he clicks that tars his data up and secure copies it somewhere (or to some cloud service) seems a better option. Quite insecure, but nothing to fall out of your pocket.

Silliness abounds

[sortius_nod]

I really don't want to come across as a troll, but this is just stupid. 15 years? Are you both mad?

Expecting this kind of lifetime from a system, without getting some sort of 15 year support contract, is just stupid. If it's 2 computers on a network you aren't going for a pricey option. I'm going to hazard a guess that they are networked via 10base2 (coax) and probably uses IPX/SPX rather than TCP/IP.

In other words, your father needs to get with the program, dump the crappy setup, convert his DB to something realistic, and get a clue when it comes to the lifetime of systems. It's going to be more expensive to upgrade now because he hasn't implemented a decent upgrade cycle (every 3-5 years is probably as long as I'd leave it).

He has left it so long that even the concept of migrating the data to a new system will be painful if not impossible.

I know this all seems trolly, but someone needs a big wake up call. Sure it's not a great idea to be on the bleeding edge of technology all the time, but 15 years of use is unrealistic and downright immature to suggest you want it.

Look at what he's using...

[Burnova]

I think most of us are getting caught up on what WE judge as "lasting 15 years." Look at the current setup, one that he's only now considering upgrading. You could probably get him a decent system now, and with proper maintenance keep it operational for as long as he is willing to use it. He obviously doesn't have too high a standard.

Use a good UPS unit with AVR on it.

[CFD339]

One of the most deadly things to longevity of hard disks and power supplies is poor power.

Using a battery backup that does AVR (automatic voltage regulation) will provide a sweet 60hz 120v (assuming you're in the US) current all the time, and will prove much less likely to fail hard disk motors and power supplies.

My experience with this stuff is over decades, and even cheapo equipment that other people see high failure rates on lasts much longer in this controlled environment.

The second thing to consider is a case designed to filter dust yet provide high amounts of air flow. You need to be able to keep the internals cool while not coating everything with dust (or in your father's case -- with pet hair and dander). Coating the vanes of cpu coolers and power supply heat sink fins will have a drastic long term impact.

A few questions to answer before deciding...

[Dahamma]

My father is also a veterinarian with a private practice... I don't know enough about the exact details of his software but can give you the high level, as well as issues he has had, etc.

First, he has gone through a few (2 or 3 not sure) completely different systems (hardware and software) in the last ~20 years of having a "computerized" practice.

When they got the first system the practice was much smaller - 3 vets in the partnership and a handful of employees. Over time it has grown to employ another 3 full time vets and a much larger staff. So that's question 1: it may be small now, but do you expect it to grow? 2 networked workstations won't be enough if he may have 20+ employees in the future, and deciding on something today (hardware and software) that at least supports upgrades will go a long way to prevent having to redo the whole thing later.

Question 2 is related to the nature of his practice. Is it a relatively low-tech, rural practice or is he planning on modernizing/keeping up with technology? Back in the 70's the most high-tech equipment in most practices was the x-ray machine. Since then, my dad's practice has added an ultrasound, laparoscope, and most recently a digital x-ray that allows inexpensive, near instant access to results (without having to develop, etc) as well as convenient storage, display on a number of terminals in exam rooms, even convenient consults from remote specialists. That's in addition to all of the other benefits that come with professional veterinary software packages, like integration with outside labs to get faster test results, tracking of inventory and reordering, etc.

Question 3: how much does he care about his data/computer systems? If down is it a minor inconvenience or a crippling liability? If the latter, do you really want to build something for him with off the shelf parts with no support? Are you available for 24/7 support if something goes wrong? My dad's practice has 24/7 1 hour business support (from IBM? or something similar). If a system goes down, HDD dies, network is flaky, etc they will have someone there in less than an hour to replace hardware, diagnose issues, restore backups, etc. Sure, that service costs money but has been necessary several times over the past couple decades and saved their ass when it happened. On the other hand, if your father is basically using the machines for payroll, inventory, and bookkeeping, he might be ok with a simple backup system and your help when something goes wrong...

Anyway, I know my dad's practice now has a central server (I think just standard workstation HW with RAID and nightly backups?), a few terminals (I believe all Windows-based, since that's what the veterinary SW runs on), and most recently a medical grade monitor and high-res video card for x-ray display, along with a couple of WiFi laptops they use in exam rooms to show x-rays, look up histories, data entry, etc. All of it comes with 24/7 HW and SW support, which for their type of usage (and the fact they don't want or need a full time IT employee) I'd consider a must have...

Anyway, hope that helped. But to summarize I'd rank the goals as (not counting cost, which of course needs to be factored in depending on personal situation):

1) minimize downtime/lost revenue
2) allow modernization/support for new technologies as necessary
3) scalable if/when the practice grows in the future

What I would most definitely NOT worry about is the latest fancy hardware. If he's still surviving on a 486 with 8MB RAM, any reasonable modern HW will be cheap and more than enough. By all means go for reliability over performance, especially if you are doing it yourself. If buying HW/SW/support from a professional company, they will make sure the HW is adequate and reliable (since it costs THEM much more in the long run if it isn't).

Re:A few questions to answer before deciding...

[Anonymous Coward]

My father too is a veterinarian. He is in a multivet practice with 30 computers on its network. From what I understand, in the veterinary world, you can get your software/hardware from the same vendor (such as ImproMed or Infinity [I think]). They will provide support and help you convert your old data. It is a bit pricier than building the machines yourself, but if your father is not too tech saavy, it might be helpful.

Re:A few questions to answer before deciding...

[Anonymous Coward]

Yep, you guessed it, my father is also a veterinarian with a small private practice. I handle all his IT stuff. There is something about veterinarians and being slightly "luditish." Get him to upgrade. You will be amazed at what the practice managment software that is out there today( we use avimark, tried intervet) can do. Both those companies have database conversion options to ease your transit. Get yourself a simple network setup with some reliable business class machines and servers, and enjoy all the stuff your new practice software can do.

Re:A few questions to answer before deciding...

[Anonymous Coward]

If all that was wrong with the old system was the hardware getting to old replace the hardware and run the old dos apps from a hypervisor or dosbox/vmware.

This lets you have the new system in a form where it can be moved to new hardware without fuss in the event of hardware falure.

There is no point reneginering something just for the sake of it. staying with DOS based apps is a a great way to avoid viruses... Conficer does not know what to do with a win95 box :)

My $0.02 on hardware
The little atom boards from intel are great, cheep, very reliable and multi cored. They make exelent soho servers and run umbuntu, samba & vmware like a champ.

Personaly I'd Grab a server quality hdd (raptor or something made for video survalence). They should last a bit longer :)

Lastly engineer out all fans (even get a passivly cooled psu.

One last thing, use some OSH relevent to medical areas:
e.g. us a rubber (roll up) keyboard to avoid spills damaging worstation. Also they are more easly sanitized.

Server SATA drives vs desktop SATA drives?

[ivoras]

Hard drive manufacturers usually have two product lines - "enterprise" SATA and "desktop" SATA. I've almost always bought server drives for servers and desktop drives for desktops but I've always wondered - what is the difference? The few "desktop" drives I misused in servers seem to work as reliably.

Pricewise, the server drives of the same capacity are more than twice expensive than desktop drives. I've heard anecdotal information that the desktop drives perform slower in RAID arrays, but nothing backed with numbers under controlled environments.

Since the warranty periods are equal (!!), where's the difference? Is the higher price on "server" hardware just spent on the brand name? I recognize that the warranty periods are slightly differently defined - the "desktop" drives are rated at some combination of "x hours per day" and the server ones are supposed to be used 24/7, but again, does it matter?

Underclock

[wurp]

I don't know anything in particular about getting systems to run for a long time, but when I saw your request I thought of a couple of things that might be helpful:

1) underclock the CPU. Then it runs much cooler, and if the fan goes out, it may run just fine without it if you underclock enough.
2) low end, underclocked graphics card - see above
3) flash instead of hard drive. My biggest problem has always been the HD going out. It sounds as if your dad's needs are light enough that you could get enough flash memory cheap. No moving parts = much longer lifespan.

Re:Underclock

[Pentium100]

So, get a PC with Intel Atom CPU and an integrated VGA. This, of course, applies only if you do not have a lot of PCI or ISA cards that you need to use. If need to use an ISA card, then your choices are limited.

I don't know about flash vs hdd reliability, but those USB flash drives go bad all the time (not for me though - I don't use one :))

Space Shuttle

[Anonymous Coward]

I hear the space shuttle is going to be scrapped soon, maybe some of the old gear from that will be hardned enough to last 10-15 years use as a PC.

Monitors

[Anonymous Coward]

Make sure you grab an old style CRT quick.

The manufacturers of some LCD screens are intentionally building in failure by putting powerboard with capacitors up against a metal backplate. Because of the heat from the power, the capacitors dry out in about 3 years, right as the warranty is going. I've got 3 CRTs over 15 years old which still work fine and several LCDs I have had died between the 3-5 year mark.

9 years on a Celeron 400 rig

[gatkinso]

Damn thing did 9 tax retuns. A power spike killed it - ran in wall power the entire time.

It wasn't the fastest rig, but it was fine for stuff like taxes email, and some nondemanding games of the day.

Built it myself. Asus. Celeron 400. Several drives over the years.

Re:Most 15-year old Sun workstations are still use

[Anonymous Coward]

Take an Ultra 1 or Ultra 2 - they are still rather useful computers, and the OS they run is rock solid by any standards.

Uhm... it all depends. We've relegated a pile (and by "a pile" I mean > 25) of Ultra-1 and Ultra-2 Suns to the scrapheap -- and we did that because they were used up, not because we had no use for them.

Use scenarios are at least as important as component specification when determining longevity.

Virtualize?

[Anonymous Coward]

Hmm, well, he seems to be fine with Windows 95....

Why not just virtualize on some small low end server from Dell. Then, use thin clients to connect.

You could upgrade the OS, maybe... or not, whatever you like... but this is likely the only way you can upgrade the hardware and keep everything intact.

Reminds me of a story I once heard...

[Capt. Skinny]

Once upon a time, in a kingdom not far from here, a king summoned two of his advisors for a test. He showed them both a shiny metal box with two slots in the top, a control knob, and a lever. "What do you think this is?"

One advisor, an engineer, answered first. "It is a toaster," he said. The king asked, "How would you design an embedded computer for it?" The engineer replied, "Using a four-bit microcontroller, I would write a simple program that reads the darkness knob and quantizes its position to one of 16 shades of darkness, from snow white to coal black. The program would use that darkness level as the index to a 16-element table of initial timer values. Then it would turn on the heating elements and start the timer with the initial value selected from the table. At the end of the time delay, it would turn off the heat and pop up the toast. Come back next week, and I'll show you a working prototype."

The second advisor, a computer scientist, immediately recognized the danger of such short-sighted thinking. He said, "Toasters don't just turn bread into toast, they are also used to warm frozen waffles. What you see before you is really a breakfast food cooker. As the subjects of your kingdom become more sophisticated, they will demand more capabilities. They will need a breakfast food cooker that can also cook sausage, fry bacon, and make scrambled eggs. A toaster that only makes toast will soon be obsolete. If we don't look to the future, we will have to completely redesign the toaster in just a few years."

"With this in mind, we can formulate a more intelligent solution to the problem. First, create a class of breakfast foods. Specialize this class into subclasses: grains, pork, and poultry. The specialization process should be repeated with grains divided into toast, muffins, pancakes, and waffles; pork divided into sausage, links, and bacon; and poultry divided into scrambled eggs, hard-boiled eggs, poached eggs, fried eggs, and various omelet classes."

"The ham and cheese omelet class is worth special attention because it must inherit characteristics from the pork, dairy, and poultry classes. Thus, we see that the problem cannot be properly solved without multiple inheritance. At run time, the program must create the proper object and send a message to the object that says, 'Cook yourself.' The semantics of this message depend, of course, on the kind of object, so they have a different meaning to a piece of toast than to scrambled eggs."

"Reviewing the process so far, we see that the analysis phase has revealed that the primary requirement is to cook any kind of breakfast food. In the design phase, we have discovered some derived requirements. Specifically, we need an object-oriented language with multiple inheritance. Of course, users don't want the eggs to get cold while the bacon is frying, so concurrent processing is required, too."

"We must not forget the user interface. The lever that lowers the food lacks versatility, and the darkness knob is confusing. Users won't buy the product unless it has a user-friendly, graphical interface. When the breakfast cooker is plugged in, users should see a cowboy boot on the screen. Users click on it, and the message 'Booting UNIX v. 8.3' appears on the screen. (UNIX 8.3 should be out by the time the product gets to the market.) Users can pull down a menu and click on the foods they want to cook."

"Having made the wise decision of specifying the software first in the design phase, all that remains is to pick an adequate hardware platform for the implementation phase. An Intel 80486 with 16MB of memory, a 300MB hard disk, and a SVGA monitor should be sufficient. If you select a multitasking, object oriented language that supports multiple inheritance and has a built-in GUI, writing the program will be a snap. (Imagine the difficulty we would have had if we had foolishly allowed a hardware-first design strategy to lock us into a four-bit microcontroller!)."

The king had the computer scientist thrown in the moat, and they all lived happily ever after.

How to build it? The same way as any other...

[geekmux]

How do you build a system that a BUSINESS is reliant upon? Easy, the same way you build one for a multimillion dollar company.

Weigh the risks of the "system" going down hard. What impact does it have on the business? If he can be down for a few DAYS, then no big deal. If he can only be down for a few hours, then the hardware obviously needs to be a bit more reliable (at least throw some RAID in there).

Regarding the 10-15 year "wish", throw that out. Sorry, but you probably paid $3000 back in "the day" for that state o' the art 486, and these days, you can afford to replace said hardware 3 or 4 times over for that price. Purchase what is needed based on business impact, and simply plan on replacing every 5 years.

Probably not

[Rix]

But if it's been run for 15 years, it's probably picked up every one it is susceptible to if it's connected to the net.

One thing about VMs... consider backups

[mlts]

As others have stated, running a VM is a wise choice as well as buying server grade components for maximum life.

I would add one addition though: Make sure to factor in backups as part of the system. The software that the original poster is using probably precludes backup software inside the VM, so one will need to backup the VM image files. I know VMWare can separate the files into 2 gigabyte chunks that can then be backed up to a volume using Backup Exec or Retrospect. This way, backups can be taken of the VM every night, and a restore is as simple as dumping the VM disk files back into the original directory.

I would recommend the backup volume be either an internal array doing RAID 1 or RAID 5/6, or an external RAID device such as a MyBook Mirror Edition or a Drobo that handles the "heavy lifting" of the RAID management for the computer and appearing as a single drive. This way, the backup data is protected from a single drive failure.

Finally, I'd also recommend periodically attaching an external hard disk, copying the backup data from the RAID volume, then unplugging the external hard disk and storing it offsite (perhaps an Iron Mountain tub.) Multiple ones would be best so you can do a grandfather-father-son rotation. Of course, don't forget to back up the backup software and its license keys (usually tossing it on a couple USB flash drives will do the trick.) If the VM image is small enough, you might be able to back it up to a USB flash drive, although I highly recommend backing the whole machine up, OS, applications, and the virtual machine.

Even though some hardware can last 15 years, I would not bet your data (and your business) on it, and highly recommend a two-tier backup solution so you can get back up and running if something does happen. Even the best servers will fail, which is why businesses that need 99.999% uptime go with clustering, high availibility, and multiple sites.

Free DOS on a fanless VIA Mini-ATX

[Hadlock]

Seriously if it's a DOS program, just run it on Free DOS, off of either flash memory, or a 5400rpm drive, doesnt really matter. Or run a regular drive that backs up to flash once a week or something. Plugged into a fanless VIA mini-atx motherboard you can just mount in the old box or whatever.

The computer, or the total investment ?

[Joebert]

Does he want the computer, or the totla investment to last 10-15 years ?

My dad had a computer similar to the one described that he paid $2,200 for. Assuming this guy did the same and is now thinking it's going to cost another $2,000+, you could probably get a decent business machine with no OS for something like $500 then put Ubuntu on it to last at least 5 years, and place the other $1,500 in a bank CD or something.

One approach

[Anonymous Coward]

The Wyse line of products make great PC's for this type of environment.

If you're keeping the current practice management package in place then turn the server into a VM. If you have the client installer then install it on the Wyse terminal.

If you don't then turn the clients into VM's running on the VMware server platform. Then VNC to connect to them from the Wyse clients.

it's a resource, not a system

[roc97007]

Look at it this way -- what you're really providing is a resource, not a system. He has some application he wants to be able to run for the next 15 years. What it runs on is immaterial.

Firstly that his current machines have kept running for 15 years without major hardware failure is a miracle. I don't think you can count on that.

I'm assuming that there's no way for him to run the app "online". The advantage of this is that it makes the workstations absolutely generic -- they just need a browser and internet access. A failure occurs, you buy a generic replacement. It wouldn't hurt to check with the company that makes the application he's using, see if there's a web-based version he can buy as a service.

If there's no other solution but to provide the service on local machines, the thing to concentrate on is redundancy, easy replacement, and backups. Hardware just doesn't last forever. Buy a drive that's supposed to last 5 years and it might crap out tomorrow, or it might last 15 years. There's no way of knowing.

But since it's a service you're providing, you can design the systems so that hardware failure doesn't prevent him from doing business.

For instance, I was at Fry's the other weekend, and was surprised at how cheap hot-swappable drive trays have gotten. For about $40 you can put two hot-swappable trays in the 5 1/4 inch bays of a standard tower. Keep a spare on the shelf, and teach him how to pull a drive that's gone red and put in the spare drive. Then take the broken drive plus tray to someone who knows one end of a screwdriver from the other for replacement.

Get a case that can be opened without tools.

Case fans are easy to replace -- provide spares. Buy a gamer's case that takes extra case fans and put them all in, so after one or two failures he'll still have cooling. Buy high quality ball bearing fans.

Get him a small cheap canister vacuum cleaner and put "vacuum out computers" on his schedule twice a year. Include written directions on how to power down, open the case, and gently vacuum out the dust bunnies.

Put Windows on automatic update. (I hate to say it, but for non-computer people, it's probably best.)

Don't buy a subscription-based antivirus. He'll forget to renew it, or won't know how, and then he'll get infected. I use AVG Free because I don't have to worry about it expiring.

You need to do something about backups. These days I ghost to a third drive once a month and put it in a safe place. Your mileage may vary. If it's too complicated, he probably won't do it.

That's all I can think of at the moment. Good luck.

Perfect...

[hurfy]

I have a set of 4 matched Pentium 3 450's i need to dump :)

Seriously something like that should work. Need 2 workstations get 3 working and parts, if they are still running odds are as good as new stuff :(

Still using a P3-1000 and a travan-3 tape drive at work ourselves.

Easy: use server parts

[dramaley]

Build it yourself with "server grade" components. My previous machine was almost 10 years old when i retired it. It had 0 hardware failures in that time, and i retired it only because i wanted something with better performance. The previous machine was a SuperMicro motherboard with dual Pentium II-450, 1/2 GB ECC RAM, and SCSI hard disk.

I'm sure my previous machine would have lasted another 5 years if i'd really wanted to run it for 15 years. The other thing i do is about once per year shut the machine down, open it up, and clean out all the dust. At that time i also check that the fans are all functional and replace any that aren't. Really, though, as long as you don't let fans get overly clogged with dust and you don't buy the cheapest ones they are fairly reliable and will usually run for years. If the machine is placed on the floor, it should probably be cleaned more than once per year as it will suck in more dust than if it is placed higher.

My current machine was built with a similar philosophy as to the previous one, and i expect it also to last about 10 years before it is too slow for my taste. Again i used a SuperMicro motherboard and ECC RAM, but this time with SATA instead of SCSI disk. However, i chose one of the "server" SATA disks that is basically disk built to the same standards usually associated with SCSI, but with an SATA interface. That required giving up some capacity (it is only 74 GB), but it is large enough for my needs since i'm not storing multimedia on it. I added the additional requirement of wanting my current machine to be quiet, so i replaced the heat sinks on motherboard and CPU with larger versions, and then used large fans that spin slowly.

Buy an off the shelf system

[Fuyu]

Buy an off the shelf system with a 3 year warranty for ~$500 (the Dell Inspiron 530 fits the bill - http://www.dell.com/content/products/productdetails.aspx/inspndt_530?c=us&cs=19&l=en&ref=dthp&s=dhs). As long as he continues to run the same application, there is no need to upgrade the hardware. He'll be able to buy 5 systems to last 15 years for the price of a system that's guaranteed to last 15 years (is there any such thing?). Using virtualization software will future proof the system by allowing the system to run on any hardware/OS that supports the virtualization software.

Re:Vet Software

[TaoPhoenix]

Yes, you're on the right track.

Come on gang, $300 in hardware is peanuts.

The guy "likes what worked" but he needs an exit strategy for his data. I know how tough retraining can be too, but he at least needs the info while his "refreshed system" is being offered so he can be informed. Both I and my coworker "didn't need fancy cell phones" until one random month apiece we "just grew into them", and then became more productive.

DOS software is pretty aged now. I really shudder to think about it ANOTHER 15 years from now. I don't know where the golden point is in Vet software, but it should be looked at to provide the decision context.

Risking Flamage

[Anonymous Coward]

At the risk of drawing massive flaming, have you looked into eComstation? It's (gasp) OS/2 with a journalled file system. It runs on minimal hardware and actually is supported. Slap it onto a couple P4 single cores with SCSI Raid 1 and have them back up to a NAS box. Looking at about two to three grand for the whole setup. Mind you, I'm not saying you should do it, just look into it. It might be an option. Otherwise, the same hardware will run almost any good Linux Distro and also XP Pro. There, I've said it. Flame away.

Thin Client

[Anonymous Coward]

You could by rights run a thin client setup with a remotely hosted redundant vm setup, that could run over a pair of redundant internet links. (or even on a server in the back office)

Then you have the ability to change whatever hardware you see fit remotely, and make it effectively invisible to the end user.

Also by running a thin client front end machine, you are effectively getting rid of pretty much everything that makes a desktop machine fail, those being moving parts, fans, hard-drives, heat is another factor...

I know those SunRay clients have been known to run a very long time in very harsh, dirty, grimey conditions.

As another son of a veterinarian....

[Dahamma]

Ok since I realized I spent 15 minutes trying to provide practical info to an anonymous submitter who will never read it, let's see if I can game the system by getting mine higher in the list - replying to a random post at the top to link to a sadly lost post at the bottom ;) Why bother? No idea, I guess I just felt like as a tech geek who has talked to his vet father about the SAME issues, I could provide something other than snark and cynicism to the / community for once...

http://ask.slashdot.org/comments.pl?sid=1188605&cid=27468991

Re:As another son of a veterinarian....

[Anonymous Coward]

Hi dahamma, i am the anonymous submitter, thanks for the link to your more elaborate post!

to answer a few of your questions:

1) no i dont expect the practice to grow, my dad is a mostly domestic animals vet in a small town (25k people), with so much competition in the area, that growth beyond 1 full time vet isnt really realistic, only a merger between two local practices would bring such growth about, in which case the computer systems most likely need a rethink anyway. My father has been working this way for ~25 years now, and he hopes to retire in ~15 years, hence the 15 year lifespan if possible

2) his practice is very low-tech in terms of equipment compared to what your dad has. No digital x-ray, no echoscopy or such, but this off course is partially because the current IT infrastructure doesnt support it. My dad isnt a technofobe or anything, but upgrading both the computers and x-ray machine when both are stil functioning correctly didnt appeal to him

3) Downtime is off course an issue, over the past 15 years i'd go so far as to claim that he at most had 1 incident per year tops. Now when something happens which takes down the system my father still remembers enough from his non-computer time to keep the business going for a few days by hand, but all financial/medical administration is in the system, so data loss isnt acceptable

As for support, on the software side he has *some* support from the software vendor, but hardware wise, my father and i have pretty much been the only support for the past 10 years, my dad did a few fan-swaps, and i took over when i started computer-hobbying.

As for reliability V performance, i wasnt even considering performance to be any sort of priority, in 15 years it wont matter if he has a 15 year old sempron of a 15 year old quad-core, both will be hopelessly outdated, yet still more then enough for the workload.

I dont consider a server to be all that interesting, the main front-desk system serves as the database server for the work-room system, which should be enough, so i dont see the point of a dedicated server.

As for support, i realize that a pro-firm with a support contract will try to have their ducks in a row reliability wise, but is there any firm that is willing to attempt a 15 year haul on any system? In 10 years anyone who helped set up the system will most likely be gone, and how many companies will be willing to keep spare parts in stock for 15 years without asking for a bajilion euros?

anyway, Thanks Dahamma, loved your post!

How about laptops?

[l33td00d42]

I haven't heard anyone mention laptops. Well-built laptops such as ThinkPads are probably top-to-bottom the most reliable hardware out there, especially if you get a low-power, preferably fanless model. Plug in a USB keyboard & mouse and probably an external monitor and go. I'll defer to others on software and hard drive suggestions, but will add the anecdote that the only laptop hard drive i have had fail was a low-end SSD, which had complete data loss after only a few months. When i dropped my open, running ThinkPad T61 about 5 feet onto a wooden floor, the mechanical drive only lost about a 5GB region, which i have allocated around and had no problems since. Barely even cosmetic damage to the laptop.

I'm kinda in the vet biz

[PJ1216]

Sorta. I write veterinary software for a digital x-ray machine. If he takes x-rays at all, he should soon be starting to look into going digital. If it really is a small practice, it may not be worth the cost right now (if he's taking at least a few a week, it'd at least be worth figuring out whether it'd be worth it), but traditional x-ray isn't going to be around for long.

In any case, the reason I bring this up is that it may not be a great idea to try to plan 15 years in advance. With the equipment changing, he may have to add a workstation into his network and if you choose the wrong OS, you may run into added difficulties. As far as I know, all the manufacturers out there only right their software for one OS (and most of them are Windows. I wanna try to get one out their for Mac at least, but I'm the only developer, so its kinda difficult). But in any case, a majority of them use DICOM and you'd need to make sure there's a PACS out there that has an interface you like on the OS you end up going with.

Basically, I'm just saying, there are changes in the vet field that sometimes will revamp a certain portion of the practice. Digital x-rays are just one example of how an upgrade will have an effect on your existing network.

Re:I'm kinda in the vet biz

[morganew]

My wife is a Veterinarian, and her practice has gone digital for all radiology. But the office and all radiology is PC only whereas our house is Mac (with some Linux server stuff - not important here).

I'd love to know what vendors are considering mac products for anything in the vet field. Ideally, my wife would love to be able to look at film on the mac, or even better on her iPhone.

Plums is now on the iPhone, as are a few other apps, looking at film would be the bees knees.

Re:I'm kinda in the vet biz

[PJ1216]

I would say to look into the program OsiriX (I believe thats what its called.). Its for the Mac and if I recall correctly, its also free. I know it handles DICOM and it likely can also do DICOM Send/Receive/Retrieve. This would allow you to send your files to the Mac using the DICOM protocol, so it doesn't matter what OS each computer is using, just as long as there is a way for them to talk to each other.

Drive ghosting

[cheros]

Quickie on ghosting the drive: set him up with a copy of Acronis True Image Home. I'm not sure if you can shorten the process to "click on an icon" simple, but it is by far the best software out there (IMHO) for both restore-from-bare-metal and file recovery. Just one tip: make sure you also store an ISO or the recovery start disk on the external drive you use. Saves time later. It allows you to recover a machine really from the ground up without having to re-enter a gazillion license codes. As a matter of fact, the use of Open Source stuff such as OpenOffice will probably help there too :-)

The Acronis license policy is excellent as well: you can always grab a copy off the Net and use it for 30 days to recover - so when you have just landed in the sh*t you don't need to rescue that first - good thinking. Worth the money - I have all my machines licensed for it. I just hope their Linux support improves so I have one solution for dual boots..

As for longetivity: hardware is NOT your problem. IMHO your main worry is the incessant OS updates and patches put out by MS, but at least the infernal driver search has lessened. If your software runs on generic Windows XP you ought to be OK, but keep a separate copy of the OS around. If you use Acronis as described, do yourself a favour and image the machine bare, before you install the app. That way you can reconstruct the base machine very quickly as well. In that context I like the VMWare suggestion in one of the comments - that really isolates you from hardware issues with upgrades.

Good luck.

Re:Drive ghosting

[roc97007]

I was thinking about suggesting VMWare's free bare-metal virtualizer as the best way to go virtual on inexpensive equipment, but you need a PC to manage the virtual instances, which raises a chicken-and-egg problem.

flash drive lifetime

[joshamania]

I use a CNC plasma table at work, the controller of which uses a 2gb flash drive for its main hard drive. It was manufactured sometime in 2000 or 2001, run about 8 hours a day or so...still working just fine. Not 15 years, but getting on towards 10, with *old* flash technology. I'd say flash is a very good idea for long life.

Old hardware is expensive; old software impossible

[mkcmkc]

This is dead on. NASA and the DoD pay serious money to be able to run 15-year-old hardware and software--unless you're made of money, you don't want to be doing this.

Best suggestion is to use the most Open Source software and commodity hardware that you can. Your proprietary software vendor may not be around in 14 years, and even if they are, they may no longer offer the software you need to replace/fix/etc.

[Yes, this means you, Microsoft. I designed my enterprise accounting system to run on Bob and I've been hearing about it ever since...]

Build him a $200 MSI Wind nettop and a full spare

[George_Ou]

$150 gets you a super low power device that hovers around 24 watts.
http://www.newegg.com/Product/Product.aspx?Item=N82E16856167032

Add a cheap 3.5" hard drive and a 1 GB DIMM for $50 and you got yourself a killer computer for $200 that well exceeds his previous hardware.

If you are worried about the thing failing, buy him a spare system for another $200 and load an exact mirror image on the system with all the software ready to go. Backup all the data to the SD flash memory slot and just move that SD card to the spare if you need to use it.

You've used a little more than $400 and you're saving him money on power costs.

everyone keeps talking about no fans but

[gemtech]

the best computer I ever had had (10 years of continuous operation, never shut it off, I shelved it because it was win98 but pulled it out a couple of months ago for a DOS application and it worked perfectly): extra big fan and a no frills video board (no fan, less heat). Heat is the enemy here. And setup to shut down the hard drives after 1 hour of inactivity.
.
But I've also had my share of power supply fans failing on cheap machines.

A few comments.

[drolli]

The things i have seen failing are HDs, Power supplys (heat because of jammed fans), and cheap capacitors (on not-so-cheap mainboards), and monitors.

1) Keep the power low, so ventilation and heat problem are no issues

2) Use SSDs (keep the power low, no reason they fail)

3) Use an RAID of SSDs (they are not out long enough to know how often they fail practically)

4) Buy a few more HDs/SSDs of hte same type, just in case

5) If you don't manage to build a system without fans, dust will be the biggest problem. Keeping the place clean can help.

6) Even risking being modded down: If DOS did the job the last 15 Years, think about Freedos. Or DOSEMU running on to of a linux kernel.

7) Buy a high quality power supply and and mainboard (not a very new one).

8) Make a Virtual Workstation.

Migration options?

[alc6379]

There are plenty of options out there for his particular circumstance, like AviMark:

http://www.avimark.net/

You can scale that up, and it runs on commodity hardware. Honestly, he's lucky that he got that setup to run so long... You could take a software package like what I just mentioned and put it on darn near anything that you want to run it on, older or newer. How many people in ANY field expect something like that to last 15 years?

Lower power == longer life (sorta)

[Theovon]

Slashdot doesn't give me room for the title I wanted to enter, so I have to write something lame.

Anyhow, I would suggest that you buy something that is (a) not on the cutting edge, and (b) designed for ultra low power. The reason is that the newer the silicon, the greater the impact of aging effects. Process variation and the effects of heat and just general usage have a greater impact on the silicon and interconnects in smaller transistors. With low-power devices, heat and power are minimised, further reducing the effect of "wear" on the device. It'll remain reliable for a longer time.

Others have mentioned moving parts. See if you can get some Via or Atom system that you can run with passive cooling.

The one place where I might suggest that you splurge on the latest tech is the hard disk. You might want to consider an SLC flash drive. It'll cost more than the whole rest of the computer, but with no moving parts, it'll last a damn long time. And don't worry about the effects of write cycles to the SSD; today's wear-leveling will make the drive last AT LEAST 10 year, but probably much longer.

You might consider doing your short-term backups to an external USB hard drive. Since the backup drive is only going to spin up when accessed for backups (if you set it up right), then it won't get much use. Do your medium, and long-term backups to DVDs and/or tape.

I had the same issue with my dad's office

[Anonymous Coward]

My solution:

* Buy a mac (kept it cheap by buying the lowest end mac mini)
* Upgrade the ram on the mac mini to the max possible
* install Parallels, and use Parallels to create an image of the current machine as is.
* run the current programs through Parallels
* backup the win95 image periodically

I did this a couple years ago, and things are working smoother than ever. With Leopard's Time Machine, the backup can be automated. With the OS image, Time Machine, and MacOSX's migration tool, you can upgrade the mac and migrate all user profile data over, thereby relieving you of the "15 year" requirement.

Parts List

[Anonymous Coward]

Make sure you consider the placement of the machine to minize dust/pet dander from entering the machine.

Below is a working parts list for what you need

Case + PSU: Rosewill R363-M-BK Micro Black Ultra High Gloss Finished ATX Computer Case with 400W ATX 2.2 12V Power Supply

Mobo: ASUS M3N78-VM AM2+/AM2 NVIDIA GeForce 8200 HDMI Micro ATX AMD

CPU: AMD Athlon 64 X2 7750 Kuma 2.7GHz 2 x 512KB L2 Cache 2MB L3 Cache Socket AM2+ 95W Dual-Core

RAM(x2): Kingston 2GB 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400)

HD: Patriot PE32GS25SSDR 2.5" 32GB SATA II Internal Solid state disk

CD: LITE-ON Black 22X DVD+R 8X DVD+RW 8X DVD+R DL 22X DVD-R 6X DVD-RW 12X DVD-RAM 16X DVD-ROM 48X CD-R 32X CD-RW 48X CD-ROM SATA 22X DVD Burner

Windows XP Professional

Veterinary Practice Manager

Office 2007 Small Business

Now you've done it ...

[gordguide]

" ... How do I provide a workstation to last 15 years? ..."

Answer #1: Apparently you should ask your dad.

Answer #2: Look what you've done: now everybody will want one.

Old is good?

[WEqR0lDRR6I]

I'd go with a fanless 350MHz Pentium II, running Windows 98SE, with 256MB ECC RAM(some DOS apps flip out if you have that much extended memory, though), on an Intel BX-chipset motherboard(Asus P2B comes to mind). Go with a name-brand power supply like Antec, non-PFC(you do want to use it with a cheap UPS without worries over compatibility, right?), with a power rating of 500W or more(not that you need it, it will probably last longer if you don't push it as close to its limits). Install two large, quiet chassis fans from different manufacturers. I'd use a 5400 RPM laptop drive(will probably need a $5-$10 mount/connector adapter), or even 7200 RPM, 20-80GB or so, with a 5-year warranty. Too large and the BIOS will have problems, and you'd probably not use it anyway.

If you want something newer(and you may, as I'm not sure what the lifespan on older slot 1 motherboards are...), you may want to look into a fanless socket 370 VIA C3, running on an Intel BX2 motherboard(only because it supports ECC, you can see other chipsets that nominally support it here: http://www.anime.net/~goemon/linux-ecc/).

Oh, and don't get a cheapo $10 case. They'll warp and bend, and could conceivably short out your motherboard. Not to mention slice open your supple flesh. ;)

Flash & oft written blocks

[omb]

It is very easy to program round this restriction, but you do need kernel awareness which AFAIK is in .30 which will be with us as STABLE 30.1+ in 10-16 weeks. Distributions 6-12 months.

The trick is to keep a smallish pool of blocks in your back pocket, and a table, usually only normally read, of where the wearing blocks are, if you use two blocks/wearing block and keep, in the first block a count of how many writes happened to the second block, you can re-vector critical writes to get as big a MTTCW (mean time to cant write) as big as you need. For example, considering just the Superblock, with 5 M back pocket space you multipy the MTTCW by 5000.

upgrade

[Anonymous Coward]

How about facing fact its time to replace the software and move it to a more modern version of the software and hardware.

I would hate to have a business running on non-supported anything.

full-o-crap

[Anonymous Coward]

didn't anybody catch the whole running win95 on 8mb of ram??? this guy is full of shit...

15 years

[Anonymous Coward]

Buy or build a machine with reliable, but not cutting edge components. You dont need a $400 video card. Just one with a reputation for reliability. Don't connect it to the internet unless you have to. Install extra fans. Open it up once a month and blow all the dust out of the fans and heat sinks. Power it with a nice heavy duty APC surge protector with battery back up. If it really HAS to last, buy duplicates of the components. Extra motherboard, RAM, etc...If, in 8 years you blow a mother board, you get the fresh one out of the closet and install it. But I think you would be better served by backing up your data to an external hard drive and replacing your new system when it eventually dies.

Maintenance will determine longevity

[Rick Genter]

It doesn't matter whether you build or buy, it doesn't matter what OS you run, what will matter the most is how well is the system maintained.

By maintained I mean:
- is it kept in a clean, well-ventilated environment?
- is it maintained periodically (opened up and dust blown out using compressed air, fans checked for proper operation, etc.)?
- is it locked down so that Ron Receptionist doesn't go install Whiz-Bang's Flashy Sudoku, Trojan and Keylogger?

Over the lifespan you're talking about, component failure due to excessive heat and/or dust-induced short circuits is probably your biggest concern. If the system is also locked down against malware, then it should just keep running...

(Owner of an 11-year-old Pentium III-based Dell box running Linux that's been his main media/CVS/DNS server for the past 7 years.)

Hey - web based!

[Anonymous Coward]

Just whack together some cheap desktops and tell him that he should not have any troubles for about least 5 years and that is about as much as you can expect from modern hardware - which by the way is exponentially cheaper than the old hardware anyway.

You can make a server yourself or you can even get a refurbished second hand server (don't laugh, I use them and they are quite reliable).

Software? Everybody is leaving this out!
http://sourceforge.net/projects/vet/

Vet software, probably does all the things they do now plus more. Web based so you could even have it professionally hosted (this also means you could give his business web presence + email!).

You can even make a cron job to do on/off-site backups :)

from this coward's point of view.

[Anonymous Coward]

First, test on a new machine (a more updated one at home maybe with some VMs) to see if the old software will work on the newer OS (sometimes Linux running Wine in Win95 mode works better than Windows XP)

Once you found the updated platform to run the code on (I hope Linux+Wine works) you can then pull in a professional group like HP/IBM/Sun etc. etc. etc. to have them help you get this running.

When it comes to making / finding hardware that'll last 15 years. I hate to agree with the rest of them, but I say good luck. Obviously the worries will be first at the moving parts, then at the power systems, then at the RAM and CPU, and finally the Mobo. (first mention = highest chance of replacing)

Slower speeds at the servers, higher speeds at the workstations. Good luck.

Lazy is as lazy does

[Anonymous Coward]

The lazy solution would be to build new PCs as the originals fail and load them with FreeDOS. Should work okay with your client's ancient app while still supporting modern hardware.

The new PCs wouldn't need much horsepower, so you could use a cool-running Via C7 or Nano, which could feasibly be run without a heatsink. (One less fan to fail) The workstations could boot from flash (CompactFlash to IDE is a cheap solution) and give the server a hard-drive with a very good manufacturer's warranty.

Just steer clear of exotic hardware and whatever replacement should last several years.

Overbuild selectively

[onyxruby]

First, let's assume your client wants the 15 years for a valid reason and so on (a consultant should always ask why the client has certain requirements). That being said you should build a computer that is more forgiving. This means keeping the computer as cool as you can with as few moving parts as you can. It wont cost as much as you think, but does require careful planning.

This means you want a larger case with room for 120mm fans, and you want at least one in the front and one in the back. Buy the highest quality fans that you can, look at the reviews for the fans and avoid ones with flashing lights. Overspec the power supply by at least 50% of the requirements, ideally 100% and get a high efficiency power supply. Again you will want to buy a high quality component that has posted positive reviews. Check out HardOCP, they have some pretty good power supply reviews. The intention is to keep the ratio of the demand load of the components as low as you can compared to the capacity you can supply.

Next you want to get an Intel brand motherboard, they aren't fancy, they aren't any good at overclocking, but they have arguably the best quality control process of any mobo manufacturer. Make sure your board has been on the market at least six months to allow the 2.0 version to be out (where they work out the bugs found in 1.0 boards). Next you want an low power CPU that runs at a cool temperature. You now have a case to underclock your CPU - you want it to last a long time. Buy a third party fan for your CPU with heat pipes and a 120mm fan, again you will want to review this on sites like newegg.com

Do the same thing for your RAM, buy quality components capable of exceeding the load (speed / voltage) and underclock your RAM. While underclocking may come across as odd to some people, it has it's place as surely as overclocking does.

For your disks, buy a paid of flash disks and mirror them for redundancy, buy a simple DVD drive that has been out for a while and is not bleeding edge. Once you have your hardware get flash happy, and upgrade to the latest reviewed stable firmware of everything. Install your OS of choice, patch everything and run a burn in for all components for at least 24-48 hours.

After everything checks out teach your father how to use an imaging program to capture and restore an image. Last thing you do is make sure you have a copy of the image offsite.

Anonymous Coward

[Anonymous Coward]

- create 2 virtual machines from current workstations
- buy 2 mac mini, one for the desk and one for the backup :>
- move the VMs to it, he can keep using his old software and use OSX for mail, web, etc

Better Planning

[Dewser]

Aside from the hardware, you also have to consider support for the software. Microsoft will limit support on the OS eventually. Windows 95 was EOL a very long time ago.

The best bet would be to plan better. Most manufacturers will not warranty workstations for more than 5 years. Shortly after the system will become End of Life. This is of course if you are going with a brand computer. If you build them, you will need to check the manufacturer warranties on all the parts. So you need to put some IT planning into affect.

If he wants something that will last long, your best bet would be to look into a small business level server. The hardware in an official server is designed to run for long period of times with longer life spans as well. Desktop grade hardware will eventually fail and it certainly will if it is running extended periods of time.

Now on top of all that, you also need to consider support for his business application. Can this database and software be migrated to newer platforms? If not then he may be looking at other problems beside just keeping his hardware running.

Hope this helps.

Don't bother

[crevistontj]

If he bought a 486-66 in 1994, he probably spent $2500+ on it. Now he can buy a system that is 20 times faster for a tenth the cost. Spending a bunch of time and energy and cash to bullet-proof a machine to last through 2024 would we a retarded waste of time.

Re:Old hardware is expensive; old software impossi

[Anonymous Coward]

Surely the insults started flying when you first contemplated designing it to run on Bob. Why did you continue?

SUMMARY

[Anonymous Coward]

You're running a DOS app on Windows 95. It does everything you need. You're afraid the hardware will fail sometime soon.

The easiest things to do is install Win95 and the DOS app in a virtual computer using software like VirtualBox, VMWare or QEMU.

You can run those apps on any modern hardware under Windows. or Linux or Mac. Pick what you are familiar with.

When you want to get fancy, you can use the tools of the host OS (Windows or Linux or Mac) to backup the datatbase and do other useful things: Windows 95 need only run the DOS application.

Yes, you could run the DOS application under Linux with WINE or FREEDOS. You could also buy a Sun server or CRAY supercomputer, but why complicate things and increase your learning (and frustration) curve?

By going this route, you can use plain old computers and plain old operating systems and keep things for your Dad just like they always have been. And you won't have to worry about this again until after he retires.

Did the original setup come with a 15 year warrant

[msimm]

..y? What this guy is trying to do *is* more difficult, if he feels he has to he should be able to explain that in terms that convey he isn't simply trying to extract more cash.

Forget the overparanoid slashdotters ;)

[cadu]

I can't guarantee that you're getting 15 years out of it, but build a cheap system with good quality parts and no moving parts at all and you're set

why don't you try something like this:

1-cheap but good quality/sturdy case, full atx, prefferably without a power supply

2-GOOD quality ATX Power Supply, rated quite a bit higher than you're going to need.

3-an (mini/micro/pico)ATX all-in-one good quality Atom based motherboard, with 1 or 2gb of DDR ram, that'll give you good performance and memory, if you have 2 memory chips you can swap them in case just ONE fails, Atom is already cool enough but put in a passive cooler over it just in case :P

4-(pick your poison) an SSD OR a CompactFlash/WhateverCheapMemoryCard-to-Ide and a BUNCH of cheap sticks, those 5 dollar 2GB [2gb wtf! bunch of years ago that was my main hard drive capacity] SD Cards you see lying aroung...they can last a looooooong time even being written at a *regular* basis, you can have an adapter plus like 10 2gb cards for real cheap, and i guarantee you that ONE card will last for much more than you're expecting, 3 or 4 years, maybe more if the system only reads stuff , and barely writes..... and, if you want to backup the database or clone the card so you have the most recent set of data you just have to pop it , plug it somewhere else and you're done.

hope my post sheds some light into your problem :)

ps: the main failure point of this system is the ATX Power Supply , since it has a fan and huge capacitors pumping and stuff converting energy 24/7 BUT as you're using quite less than the power supply is rated for, you shouldn't be facing overheat/capacitor issues.

Re:Forget the overparanoid slashdotters ;)

[Sj0]

Why use unproven Atom technology? Via has a good track record, and their EPIA tech has been fanless for a long time. Combine with a fanless PSU(Preferably mounted longitudinally so heat can rise out of it through the existing grill), a fanless case, and a compactflash with an adapter kit, and you'll have a PC that you can't hear, that should be insanely cool, that should run long past the end of the warranty periods (meaning it'll be long out of the infant mortality curve).

There are prosavage drivers for 9x, so you can even keep running the old operating system and software.

You can run multiple PSUs with some extra dremelling, but that's your business.

Here's the only way

[SpacePunk]

Get a new case, gut it out, literally. Hinge the front panel on one side so that it can be opened. Put ya some sort of document holding system in there so he can store folders in it. Viola! A 10 to 15 year computer... of sorts.

Re:Most 15-year old Sun workstations are still use

[keeboo]

I love Sun hardware, but we've had an Ultra-1 which slowly fried its processor (which is soldered directly to the PCB btw). During its last months of use we had to adapt a strong fan blowing directly to the heat dissipator, otherwise it would simply go crazy under high load.
Oh, we've had a pile of other broken Ultra-1s aswell.
Sparcstation 20 is another model prone to problems (due to bad internal refrigeration).

The most reliable (old) Suns I've encountered are the Classic, Ultra 30 and Enterprise 450.

But isn't the guy going to run Win/DOS apps? Are you suggesting him to use qemu?

Re:Simple answer, is it's time to redesign the sys

[bhtooefr]

Of course, DOS isn't dependent on hardware platform any more either - if it runs on DOSBox, it's now pretty platform independent, and can easily be moved to other systems.

Backup to online(Internet) storage

[kcramakrishna]

Things will definitely fail and we can't guarantee system availability without a detailed Reliability analysis of all parts in the PC. However it can be guaranteed that something will fail. The easiest way to get around all this is to not plan for redundancy in the desktop system but rather plan regular backup of the source and data to online storage. (Maybe twice daily). You can use Amazon's services which is quite feature rich. This way even if something fails, you can upgrade to available hardware without breaking a sweat. You will not be at the mercy of hardware/software vendors. And Oh, yes, use GNU/Linux. We are based in India and can build the system for you if you want. - www.rknowsys.com

d1g1t4l_nrg

[d1g1t4l_nrg]

Hey if you need volunteer help with the open source db being built send me a line. Have a good one and good luck with the long time pc rebuild.

For the scope discussed, irrelevant.

[Junta]

Yes, AMD has been on socket AM2-like for a 'while' relative to intel, but not too long they were in socket 754 and 939. Before that, Socket A. Before that, Slot A. And the generation before that (which *finally* gets you in the ballpark that original article would be relative to now) was socket 7.

On Video card, it's getting harder to find a servicable AGP card. While AGP was starting to become popular in that time frame, they generally won't accept the cards today (keyed for different voltage).

HDD he'd probably be fine replacing, it's not hard to find 40-pin IDE drives that would work in the controller of something that old..

PSU you actually could be fine, Baby AT had only recently died out, so ATX has been ubiquitous since then.

Fans haven't changed in incompatible ways along the way.

So in short, processors, vid card, and memory are hard to chase down as their interfaces have been very different in incompatible ways over time. With this in mind, ensure your motherboard and video card require only passive cooling, and that your CPU cooler accept a standard (80 mm probably) fan.

In shorter, don't sweat it, make sure the hard drive storage is redundant, that a sane backup strategy is in place, and don't fear replacing the whole system if things go bad down the road. If third party software is involved, keep it safely archived.

Not a fallacy..

[Junta]

First, your last point about external drives for backup, there are two distinct points being met here.

A sane backup regiment is indeed key to protect against more catastrophic hardware and software problems. It's also a pain to have to resort to that. Most people that have a 'couple USBs' to handle backup don't implement proper backup procedures, and are not at significantly greater protection from hardware failure than the much-maligned RAID-5 configuration.

RAID-1 is expensive relative to 5 (commonly 2n storage to store n data), and in the common usage, a degraded config is equally at risk as RAID-5 (replace the failed drive, and if your RAID-5 rebuild would have failed because of long-untouched sectors being bad, the mirror rebuild would fail too). Yes, you can do more drives in RAID-1 (I actually do in some special cases), but most don't bother to go to that expense). RAID-6 is more reasonable, but the base problem of long-dead unknown sections is still present.

This is why most array set up nowadays do background consistency checks with idle IO time, to detect and rewrite failures on otherwise unmanipulated parts of the disks..

RAID is about zero-downtime error tolerance/recovery without significant hassle. It isn't capable of recovering from everything, but it is a *lot* better than having to go to your backup media. Combined with increasingly popular online snapshotting capability, and the hassle of going to backup to retrieve a file for hardware or software failure becomes a relative rarity.

Re:Not a fallacy..

[Aram Fingal]

Actually, RAID 1 can be cheaper than RAID 5 when the required size is small enough. RAID 1 is easily and reliably done in free software. RAID 5 typically requires a hardware controller. I have heard of software based RAID that will do level 5 but I don't know whether it's any good and, last I checked, not free. The point is that you have to make up the cost of the controller with reduced cost of hard drives (due to getting 67% of total space from RAID 5 as opposed to 50% with RAID 1). Up to a certain point, it's cheaper to just buy bigger hard drives. Furthermore, as someone else mentioned, you have to have at least 3 drives for RAID 5 but only two for RAID 1.

You can also do nested RAID in free software (combinations of striping and mirroring). You really don't save much with RAID 5 until you need to beat the storage space of a RAID 01 (or 10) made of four of whatever size is most economical at the time. So, currently, you need to be after more than about 3 TB of storage before RAID 5 makes sense. Since the situation we are discussing here seems to have low storage requirements, I would think that RAID 1 is going to be both the cheapest and easiest solution.

Re:Old hardware is expensive; old software impossi

[sryx]

[Yes, this means you, Microsoft. I designed my enterprise accounting system to run on Bob and I've been hearing about it ever since...]

Brilliant! Why can't I mod this up as hilarious while still granting props for the insightful first part.

nasa research what nasa uses

[Anonymous Coward]

Well i would not go as far as to leave him with 95 i would probably get windows 98se at least that is assuming windows xp didnt work with his software i would test that first i would also see about an embeded system on a chip like device connecting to a main server then use a raid 1 or 5 of the lowest density drives you can they still sell drives under 100 gig

Get 4 identical systems and only set 2 up

[richardkelleher]

Order 4 simple (cheap) identical systems, but only get XP on two of them. The others are for spare parts and won't need an OS. Put the spares in a closet still in the boxes. In all likelihood, the displays will fail before the computers.

mgHz

[mapinguari]

Nah, milligram-Hertz (mgHz) is an esoteric unit, typically used colloquially to express the efficiency of a processor - the weight of the processor multiplied by its clock-rate.

By definition, it's the amount of energy it takes to oscillate a milligram at one hertz:
350 mgHz will vibrate 350 mg at 1 Hz, or 1mg at 350Hz equally.
350 mgHz may be expressed as 0.35 gHz (not to be confused with GHz).

As such, I believe that the OP has misplaced the decimal. It's more expected to see something along the lines of 3.5e+12 mgHz.

Re:mgHz

[richard.cs]

I think you're mistaken there, gram-Hertz has dimensions of mass over time. Clearly if we know the mass of a byte then we can translate this into bytes per second, which is probably the data rate of the ram or the cache or something. Since a 32 GB SD card is available with a mass of 2 grams this puts an upper limit of the mass of a byte of 6.25e-11 grams. This means 350 mgHz is at least 5.6 Gigabytes per second. Obviously since an SD card also contains plastic and magic smoke the true mass of a byte is lower and hence this is only a lower limit of the his machines capabilities.

Buy used. Buy lots.

[ResidentSourcerer]

Buy 4 used computers. NOT dells. Not to disparage dells, but they have non-standard form figure power supplies.

Two machiens replace his current machines.

The other two are parts.

Put a second drive in each running machine. Doesn't have to be big, but disks that sit, stick.

Put UPS or at least decent surge suppressors on the working boxes.

I ran a computuer lab in a rural school for 5 years. With 60 machines I lost 3-4 power supplies per year, and 4-5 monitors per year. Lost 5 monitors in 3 days. Surge.

Adding an extra cabinet fan set to vent through 3 removed expansion slot covers is a good idea.

Things to remember

[Anonymous Coward]

1) Build
2) Redundancy
3) Proven hardware, spend the money now so you won't have to spend it later.

Laptop solution with external HDD

[Anonymous Coward]

since the most problematic issue is changing the software, I say don't. and I propose a "design" solution :

two dell latitude e6500 with ssd ( dual 64Gb with raid mirror for safety ) with extended support for 3 years with FreeDOS ( by calling dell and explicitly asking for this os )
http://www.dell.com/content/products/productdetails.aspx/laptop_latitude_e6500

one western digital external hdd My Book® Studio Editionâ II ( with raid mirroring, 2x 1Gb should be enough )

connect the external hdd via eSata to one of the laptops, and since the actual data is about 500Mb and gMail's storage about 6G, you could set up an on-line backup using gmail's imap as ftp

another alternative for the external drive could be one that have wifi connection

this should answer your needs, and online back-up is highly recomended. and for installing FreeDOS, well, you format and install in the same hour ( if needed ) also, you could connect external keyboard/mouse/monitor to these laptops.

Cart before horse

[Monty_Lovering]

Your dad specifying a 15 year system life is like you telling him that you want an injection that will neuter your cat for five years.

Wanty wanty no havey.

Why on earth should your non-techy dad's computer wants be of any more relevance than a non-vet's animal healthcare wants?

Get a couple of big disk Core 2 with 4GB mem and XP, or buy with Vista and get a voucher so you can junk it for W7.

Install Open Office or MS Office.

Tell him it should last 15 years.

Fix it if it doesn't.

Virtual machines are the answer

[LostMyBeaver]

While it would be great to build a system that is dependable for 10-15 years, unless you have a huge budget for environmental testing of a large set of components and the knowledge necessary to interpret the figures that result from the research, there is no way to predict reliability of systems over that period of time. It's better to just try and buy what appears to be the best.

I recommend buying system components from a vendor like Tyan or Supermicro as opposed to Asus and Gigabyte for example. Those vendors survive on dependability. They sell the same components for years instead of months and from my experience are quite good at maintaining them.

If your dad has been using Windows all this time, then Windows is probably the best solution for the future. Any person that uses the same systems for that many years is not interested in a great deal of change. Windows Server 2008 is a great option for these types of environments. Use it with thin clients, check out http://www.chippc.com for example. They have a thin client that fits into a simple network jack.

Install the server as a virtual machine on the free VMWare ESXi, it should be a format supported for a while to come. The images are generally compatible with most other virtual machines, so 10 years from now, they should still work.

Install a tape backup and backup regularly the virtual machine. Get a SAS controller and install 5 drives. 3 for the stipe + parity, 2 for hot spares. This will allow for up to 3 full hard drive failures over the life of the machine. Since you're running a VM, I recommend using faster drives. Sounds like dad's been quite happy with 500meg drives for a while, doesn't sound like he needs too much storage. 73gig 10k or 15k drives should be sufficient.

Use overclocking RAM from Kingston or Micron(crucial). They have heatsinks and when you're not overclocking should prove highly reliable over extended periods.

If you'd prefer to avoid the hassle of building and possibly maintaining the machine, I would highly recommend purchasing an x86 server from Sun Microsystems with Windows and you can also use their thin clients with the system. Not as cheap as DIY in the short term, but is very likely the less expensive choice over a period of 10-15 years. Additionally, the components in Sun machines are the same ones they sell to the military, so the quality is quite high.

Here's how I have them last that long

[fedorowp]

I have experience building workstations and servers that last. Nearly all of the ones I've built for customers are still functional more than 10 years after first install.

Experience counts so I suggest you use a system builder with a similar track-record.

The more powerful the system, the more challenges in building it to last. Many of the items on the check-list below need to be balanced against the needs of the customer, including noise, environmental conditions, performance aspects, and frequently budget.

Check-list for Building a Computer that Lasts

• Minimize expansion hardware. Expansion slot connectors sometimes oxidize so the less plug-in hardware the better. This includes on-board video, serial-ports if needed, etc.
• Use a high-end board from a quality manufacture. High-end boards tend to have powerful CPU voltage regulators and are designed to support lots of memory, which reduces memory controller issues as the board ages. They also tend to be the boards preferred by early-adopters, which manufacturers are probably more thorough in validating. My current preference is for Asus as they have the highest end consumer boards which support ECC for AMD CPUs. Make sure not to overtighten the mounting screws.
• One or two identical memory modules. when memory modules are mismatched, or with more than two unbuffered modules, when the memory controller ages you're more likely to run into trouble. Use memory approved by the motherboard manufacture. ECC is recommended.
• A great power supply. An oversized PC Power & Cooling power-supply is the best choice for environments that can handle a fan and noise isn't an issue. That said, quiet is very important in many situations, and PC Power & Cooling's Silencer models certainly aren't silent under load. For those situations I use an oversized Zalman heat-pipe cooled power supply I install a Noctua fan into. With that setup you don't hear a sound from the cooling fan and the power supply runs extremely cool.
• Hard drive redundancy. RAID-1 or RAID-10 is the only way to go for normal systems. A quality true hardware RAID controller for Windows, and software RAID for Linux. A hot spare is recommended. When using a software RAID, if you need to be sure the machine will boot with a HD failure, use a hardware RAID for the boot volume. A rather neat low-cost way I'm doing that for the next Linux server I'm building is using an Addonics duel CF interface that has hardware RAID in it.
• Plenty of cooling with quality fans. No sleeve bearing fans, and if the speed of any fans is reduced to control noice, make sure they can start from every rotational position.
• Use quality HDs and install them correctly. For the past several years Western Digital's high-end hard drives have had a perfect track-record for me. The most important thing to remember when installing a HD is absolutely, positively, don't over-tighten the mounting screws. Plenty of clean power, good cooling, and eliminating any vibration being transfered to them is important. Mount them as low in the case as possible to help keep them cool, and leave space between drives. If you use Seagate drives, server class is a must. In the last server I build, I did a RAID-1 between an Intel X25-E SSD and mechanical HDs so all the eggs aren't in one brand/type of basket.
• Good power protection. I've never had a computer damaged by lightening plugged into a metal-case Tripp Lite surge protector. Also protect the cable, DSL, and modem connections, and any non-fiber runs that go outside the building. Make sure you protect all network equipment too. Plug an APC Smart-UPS into the Tripp Lite and you have total protection. No other brand or model UPS has help up as well in the long-term. Dedicated circuits are the icing on the cake, but with the Tripp Lite + APC SmartUPS combination, as long as the outlet is wired correctly, no matter how bad the power is the computer has always worked fine for me.
• P

Easy...

[Anonymous Coward]

Just buy a cheap as chips PC and some external HDD for backups. When either fails, throw it away and buy a new one.

Good luck with that

[Anonymous Coward]

I think it is harder than it used to be but I would...

1) Buy a good quality brand name fanless MB with onboard everything.

2) Use a low RPM HDD

3) Put it all in a good quality case with a modest but efficient PSU.

The general idea is to get good quality components across the board and not to drive the hell out of them. A side effect is that you will end up with a system that is also super quiet and thrifty on power.

No moving parts - or maybe VMWare

[goldcd]

Firstly - Running IT for your family is a pain. We all know you'll get nothing out of it apart from grief :) If you are feeling altruistic, then read on.
Secondly. Moving parts break. Heat breaks things. Fans sucking dog hairs into the system will break things. Assuming he's going to be using the same MS DOS app for the next decade or so, he does not need a powerful machine (which is handy). You just want some ultra-low power system (Atom? - how about one of those Asus desktop EEPC thingies - fit a cheap SSD if it doesn't come with one) - ideally just get something with a CPU and a PSU that doesn't need a fan, just a heatsink.
Thirdly it will fail. It's a PC for your Dad, it's critical to his business - therefore whatever you do will screw up. Install a backup solution. I'm assuming it's not creating vast amounts of data, so just something that'll spool the new data up the ADSL/cable/modem to a NAS/PC you can get your hands on somewhere out there.
Finally - you might want to consider VMWare. Performance hit isn't a problem in his case, and wouldn't it be nice if you could restore a complete failure/screwup in minutes (Oh and allow him to run a decent OS alongside the DOS app - as a bonus). Hardware's going to be cheap, so might as well buy a spare system. Anything does wrong - you just zap the image onto the new system and he's up and running whilst you try to work out whether the old PSU shorted, or the memory just came loose. If you feel very techy, could just setup the systems to mirror and implement a hot-standby (although possibly we're moving into the realms of overkill here).

Another alternative

[goldcd]

would be to just rent a virtual machine (and let a 3rd party company deal with all the backup/hardware gubbins).
Weak point here though would be the connection to the server - so as well as primary ADSL/Cable access from his office, you'd then need maybe 3G dongle backup on his router, then a spare client machine and then.. :)

You could always use a commodore 64.

[Anonymous Coward]

How about a personal computer lasting 22 years only to be resurrected as a working web-server.
Seems to me the old girls got a bit of life left.
http://www.c64web.com a commodore 64 web-server.

Don't Buy Spares!

[supernova_hq]

If I have learned anything about computer maintenance it is the following:

• Don't buy spare parts! (except ram, it increases in value...)
• Moving parts == bad (but SSD don't last very long yet)
• Don't rely on windows for long-term projects!

Ram is the ONLY thing that appreciates over time. Don't buy spare parts for anything else. Unless it is something that will be obselete, but that's a dumb thing to buy for a long-term machine...

If you can get him off windows (or any closed-source software for that matter), DO IT! You will always have the source code to linux+gnome+firefox+apache+mysql, but windows will probably never be available for more than 10 years.

Re:Don't Buy Spares!

[Omnifarious]

I disagree that spare parts are a bad idea. I have some old ATA based systems and it's starting to be hard to find hard-drives for them. It's also hard to find AGP graphics cards nowadays too.

The way the different components in computers connect to one another has been on an accelerating curve of change for awhile. I wouldn't give 5 years for my ability to find replacement parts for my current computers. I even bought some extra hard-drives to swap out in my RAID arrays.

Re:Don't Buy Spares!

[supernova_hq]

Unless it is something that will be obselete

I guess you missed that part. If you are using a graphics card or something similar, then that would count as something that would go obsolete.

My point was that until they stop making something, hardware costs go down, FAST. It's usually cheapest to wait until the next version of an interface comes out, and then buy your spare parts.

If you graph the cost of a particular computer part, you will see it fall sharply at birth, then go back up when the next version becomes more popular and they stop making it. The trick is to buy it just as the new one comes out.

If the guy is building a computer to last 10+ years, he is going to be buying NEW technology, thus he should not (in my opinion) be buying spare parts for while.

Re:Vet Software

[julesh]

DOS software is pretty aged now. I really shudder to think about it ANOTHER 15 years from now. I don't know where the golden point is in Vet software, but it should be looked at to provide the decision context.

Legacy software isn't an issue until you need it do something it doesn't already do. But at that point, you typically need it to do it yesterday, and migration to a replacement will take months. So, it may be best to start looking for an upgrade now, preferably to something open source and/or custom, so that migration doesn't become a hassle in future (avoid vendor lock-in).

Swap the OS

[aunt edna]

There's a chance your father's PCs will need their OS reloading at some point ... because you have to rebuild (cpu/memory/motherboard gone west.)
That could present problems if by then the currently available hardware isn't useable by that version of Windows.
I suggest you try to swap to a version of Linux & run Windows in a virtual machine, using Parallels or whatever.

Best of luck!

Onboard everything, passive cooling, underclock

[Qbertino]

Get a cheap, small, fanless, onboard everything system with external UPS and underclock it. Install DOS direktly or try running the software of DosBox on a slim Linux variant in order to cover for the larger RAM space today Comps have. MiniITX sounds like a safe bet. I have an underclocked AMD box (1 GHz instead of 1.6 GHz) in my line of computers and it runs very stable and is way faster than anything from 15 years ago.

Any no-frills wallmart computer

[Kindaian]

Will do...

Then add the current crop of OS... and a software called DOSBOX.

And you are set-up...

With the advantage of using either linux/windows/macos... since DOSBOX does the magik... (and there are other DOS around)

Don't go for the newest and greatest

[jandersen]

If he wants long-term stability, he should opt for a system that isn't too much cutting edge and also keep using it as he clearly has for the last 15 years: limits his use to what is necessary to get the job done. The only concession to newer technology, in my view, should be to use SATA drives, since they are more likely to still be around in 15 years; and make sure to have to physical drives - one for the system, one for data - that has been a lifesaver for me sometimes. And despite what they all say about real men and backups, just do it, regularly.

Probably the worst mistake you can make when upgrading is to go all-out for a new, hyper-giga machine. This is one of the things that always kills public instutions; they have chugged along happily for 40 years on a hand-crafted computer with ferrite-cores and running algol, and then they get the latest in high-end servers and all is chaos, because everything is completely different from what they are used to. To get a smooth transition, it is better to simply get a bigger and better machine and the continue to run the same apps on it; this lets people get used to the new hardware, and allows them time to istall and test newer SW little by little.

Netboot

[Anonymous Coward]

The moving parts are the weakest. If the software is up to scratch, just buy a low power machine, like a via-epia based machine with no hard drive. Make a disk image with the client software, and netboot it from a linux server.

No fans or disks at the workstation. Only one computer to maintain. Oh yeah, make sure you have RAID and a good backup solution.

DOSEmu + Diskless Workstations

[kjhambrick]

If the App itself is still good enough to manage the proctice, and it is truly a DOS App, then maybe a linux terminal server + DOSEMU + diskless workstations is the way to go ?

http://www.google.com/search?source=ig&hl=en&rlz=&=&q=buy+diskless+workstations&btnG=Google+Search&aq=f

http://dosemu.sourceforge.net/

http://www.ltsp.org/

Experience

[Atheose]

As someone with experience with hardware support for a large company, I can attest to the assertion that SSD's fail pretty often. We use the HP TC4400 tablet, which has a 40gig SSD in it, and we seem to get more of them in with dead harddrives than we do the TC4200, which uses a typical SATA drive.

This may not be true for all SSD's, but it's my experience so far.

Re:Experience

[Anonymous Coward]

As someone with experience with hardware support for a large company, one would expect you to know the difference between the technology used to connect the drive to the rest of the system and the technology used to store data in the drive. IOW, I expect that those SSDs you mentioned are SATA SSDs, and that the "typical SATA drive" is in fact a SATA HDD. Geez. Sometimes it's worth getting details right, like when you claim to be an expert.

Re:Most 15-year old Sun workstations are still use

[downix]

I do, and it has 256MB. Mind you, I don't run GNOME or KDE on it, but for basic xfce and E17, it does a fine job. Also, do not use the onboard IDE for more than your /boot or you'll feel the PAIN! PCI SATA control card FTW!

nobody has mentioned monitors

[j1mmy]

your father's eyesight is going to be getting worse with age, not better. big, pretty LCD's are ridiculously cheap these days. you can find good quality 24" screens for under $300 all sorts of places. these may not last 15 years, but they're going to be so cheap when they do fail, it hardly matters.

Why not renew all ?

[Anonymous Coward]

Why do we need to focus on the hardware to last 15 years? He had a system which lasted for 15 years both hardware and software. But asking for another 15 years is asking for a solution to be valid for 30 years total. The cost of keeping it together for another 15 years may be higher then expected. He made a wise choice when he put those together so it lasted this long (with some luck apparently). I think he needs to spend that extra money that he would spend on the extra reliable hardware, to renew the system altogher to a decent one. For example web based one with an online appointment system and/or live video feed from the veterinary. All these with a decent hardware shouldn't be higher in cost comparing to super durable hardware will run for 15 years and all those efforts to keep old programs running too. If the hardware fails in somewhere it won't be too pricey or hard to find to replace (remember the recent article about google choosing regular hardware over durable server blades) and software related costs probably will be less comparing an vmware/(name your favorite virtualization solution) consultants hourly charge. In my opinion at least this kind of solution should be also offered to that person rather then giving what he thinks his only need (which is also costing more for him in the long run)

Check out Fit-PC

[Anonymous Coward]

http://fit-pc2.com/wiki/index.php?title=Main_Page

They're fanless, small, low power consumption, pretty robust and the only moving part is the spindle of the HDD, that can be resolved with an SSD (discussed in detail by other respondents).

Cleaning it

[sherriw]

Make sure you open it up periodically to blow out the dust and _animal fur_ that's going to accumulate in there. That'll be a big source of any future hardware issues.

Re:Most 15-year old Sun workstations are still use

[blind biker]

Yep, booting from a 9GB SCSI drive :)

Thanks for the info. Knowing that Solaris 10 works with just 256MB is encouraging.

Develop a WebApp

[chord.wav]

Put it online. It should last as long as you pay the hosting bill. You don't have to worry about backups, etc.
15 * 12 * ([hosting bill] + [ISP bill] + cheap PC) [total cost of hardware made to last 15 years]

Re:Develop a WebApp

[chord.wav]

That should have been:
15 * 12 * ([hosting bill] + [ISP bill] + cheap PC) < [total cost of hardware made to last 15 years]

Computers aren't what they used to be

[uncledrax]

Unfortunately, that windows 95 box is sort of on the better part of the curve.. You're going to have to explain to your father that he got pretty lucky with that, and most people wouldn't consider a computer a 'durable good'; and although 5-7 years would be considered a good life for a desktop, 15 I think most people would consider unreasonable to try and plan for; especially since his "P-66" (P-166 I think it should be) didn't require a fan or otherwise near the level of moving parts that current machines do.

I'd see what the spec is for the (new?) software he's going to be using and get the coolest/simplest/more durable design you can do to run it. If you buy good parts, you can probably get a the processor and mobo to last that long, but PSUs and Fans have shorter lives.

Also, I'm frankly a little amazed that he's had 1x 500mb drive with all his data on it for that long. I would HIGHLY recommend he implement a backup solution for his data.. it can be as simple as some online data repository he compresses and uploads his data to (GMailFS even!). Set it up automagically to run at night when no one is in the office.

Go virtual, take backups

[WindShadow]

Get a modern machine with few moving parts, install a recent Linux, and run both machines virtual. I assume these are not 7/24 machines and could be shut down for a few minutes to take a backup, which would fit on a single DVD

Example:Newegg has a Shuttle, 64 bit CPU, 1GB RAM, 80GB disk, and Linux installed, although not the one I would use. Add a 2nd hard drive so you can run RAID-1, a DVD burner, and a modern Linux like Fedora-10. When backups are made, use software ECC on the DVDs, like dvdisaster, so they are almost certain to be readable for 15 years.

The nice thing is that these units are only $200, maybe $300 with the DVD burner and 2nd disk, so you can have another on hand and ready to go. And since you are not (hopefully) network connected, security updates are not a requirement, just run.

15 years is a long time...

[joeyblades]

However, I have a Mac that is still running and is useful after 11 years. Apple computers are very reliable. I've owned more than a dozen and only had one hardware problem and as far as uptime, they can't be beat. My Mac never, ever, ever crashes... ever!

Get three Mac Minis. Use two for day-to-day use, keep the other on the same network and use it as a backup storage system. Then **IF** you ever have a hardware failure, you can just swap one out.

You won't find a more reliable solution for that kind of money.

Can your Father...

[elprez333]

... recommend a breed of dog that will last 15 years? TIA.

Don't re-invent the wheel here

[tmick7]

Buy a bottom of the line Dell. Then buy one exactly like it. Setup your Dad's system on one. Make a ghost (or similar imaging tool) image of the system once you do have it set up. Burn it to a DVD and put it with the second cheap dell on a shelf somewhere. Use the burned DVD and second Dell if you need to.

buy a mac

[Kildjean]

buy a mac mini, with alis-vet software... more info here: http://www.informavet.com/ Mac's have the most resilient hardware vs time and performance... If he needs to run his DOS only app, you can use DOSBOX http://www.dosbox.com/download.php?main=1 its very easy to script for it. If you need more help let me know, I wont charge you a cent...

Don't

[Digital Eco Freak]

I support research scientists and get similar requests. They have been using systems for 10-15 years and want to follow the same pattern, but the long delay between upgrades makes for much more violent changes when the upgrades do happen. We have to make great efforts to migrate files and workflows, and fight with them over every change.

One group decided that their system worked and they weren't going to mess around with it. Now they are scrounging ebay for backup VAX parts and scouring old spaghetti code every time anything needs to change, hoping that they don't break it all. The migration process to a new system has been painful, complicated, and expensive.

If you want a stable system, plan for incremental changes that keep up with the times. Buy reasonably priced equipment with a good backup solution, and you can just replace things as they fail or become obsolete. Instead of a really big capital outlay once, plan for a more moderate investment with recurring maintenance and upgrade costs.

Now would also be a good time to look at new commercial off the shelf software for managing a veterinary practice. It may be another significant investment, but it will (hopefully) give him long-term commercial support, and access to new features he doesn't get with his existing system.

Your father may not like that answer, but it is the only responsible one. With the ever-changing nature of computer systems, planning not to change or adapt is planning to fail.

The real issue...

[ujoronen]

While reducing moving parts and heat generating components are important, as is upgrade path, the real issue is software longevity. The next version of your OS and specific application may not last 15 years thanks to the planned obsolescence of the industry.

My plan for something like this would include an Intel Atom processor integrated motherboard with 2GB of ram The idea is to prevent swapping to keep the HDs longer. Run a 200W DC power supply with an external AC adapter. Put the whole mess in a large steel case and block off the vents. The steel case will dissipate the heat from the HD and CPU, provided your office stays below 85 degrees.

As I said before, the OS and software are the real trick. While you could run Win95 / DOS, there are other options with clear benefits. The existing database is probably either an embedded flat file or a Lotus Smart Suite / DBase3 clone. The latter will have issues with the next part of this plan, so you may need to upgrade the software or modify it to handle a modern database like MySQL.

My solution? run Linux. You can use Linux to create a virtual DOS or Windows machine using VMWare or the Open Source equivalents. Since the database may run client-server, it's possible to run the database on the host machine, while running the DOS / Win95 image in non-persistent mode making it safe from modification.

Simply back up the virtual machines once, and your database on a regular cycle for disaster recovery. If it ever goes down or you want to upgrade the OS, simply reload Linux and restore the database and your virtual machines.

Another simpler option is to use wine or DOSemu to run the software as is, but these can have some compatibility issues if you upgrade the host OS at a later date.

This plan will allow the machines to run at a walk rather than a run, reduce moving parts to a minimum, present the least learning curve for your users, and keep you running no matter what fails.

OpenSource Animal Shelter Manager

[Brian Edwards]

Animal Shelter Manager is a free computer package, covered by the terms of the GNU General Public License. It is designed to take most (if not all) of the paperwork away from running an animal shelter/sanctuary. The system can also be used to manage a veterinary surgery or clinic.
It runs on any computer platform, utilizes other open source components and does not require a modern computer to run. http://sheltermanager.sourceforge.net/home.php

You can't

[grikdog]

The industry changes too fast. No one can predict what (or whether) computers will be like in a decade and a half. The only thing likely to remain true is that Open Office.org will still run in emulation on some platform or other — software is expensive, hardware is cheap and evolving fast. So, although you may still have OOo around, your best friend may be a virtual robot on a television screen who types (in OOo as a retroid lark) whatever you dictate.

Remote System

[DarthVain]

Not sure if this is really a good idea and it is likely too expensive but perhaps an interesting case anyway:

Have all your software developed to portal technology. Pay someone to host all your stuff remotely on servers (which they maintain anyway). Buy a really low tech computer and use it like a terminal simply to access the remote server over the internet.

You pay up front for development, and after that fees to the host.

Only drawback is if you ever lose connectivity your pretty much screwed. Anyway just an idea.

VMware Server 2.x = Free..P2V workstations.

[Anonymous Coward]

P2V the existing workstations to a VMware image. Install VMware Server 2.x (Free) on a modern dual core system with a good bit of ram.

You will be able to run these very easily... This will enable you to move the Virtual Machines from anywhere to anywhere.... Take the hardware out of the picture and ye old servers shall last forever.

I would consider using ecomstation (OS/2)

[Anonymous Coward]

I would consider using ecomstation (OS/2) for this since it runs DOS apps.

Impossible

[Anonymous Coward]

I would not even attempt this unless you will be in a cave for 15 years - Look at the technology curve, and you will understand why. Change is happening at an ever faster rate! If you think your old computers are obsolete now, you have seen nothing compared to current computers 15 years from now. It will be like having an original IBM PC or Mac - useless except as conversation pieces. I have a 64K CP/M machine in my attic which is probably closer to the truth.

    5 years may be doable... just by buying carefully and being willing to upgrade components.

K-dud

[Anonymous Coward]

Ignore all of these techno-geeks. What you need to do is find a larger hard drive, say 700 MB and clone the disk to that drive and install that.

This is because your software, drivers, attached printers, parallel cables, serial modems, etc are hopelessly intertwined with the old DOS application you are running. There is a good chance things won't work well with a modern system.

Since there are no performance problems, you are just running out of hard disk space this is the remedy. If you want to make a few improvements see if you can upgrade the ram or replace the monitor with a flat screen to save space.

Debian + DosEmu

[supernatendo]

If you are going for cheap and he is not looking into changing his software then I suggest the following: 1.)Build a system based on Rock-solid hardware, preferably something that has been on the market for an unusually long period of time due to high demand and that generates the least heat and uses the least voltage possible. This will also ensure that if a component does go bad in the next 15 years spare parts will be more likely to be found in working condition. 2.)Setup a couple of Western Digital (Stay away from Seagate at all costs!!!) hard drives in RAID 1 configuration 3.)Run the system on Debian Stable 4.)Install dosemu and use his existing software 5.)Sleep easy!

Use VMware or other virtual environment

[ChrisA90278]

If his software runs under DOS then rub a virtual "DOS" system inside a modern OS

That Modern OS can be Mac OS X, or Linuix. Or even Windows.

Once the computer is running out of a VM yu no longer have to worry about the hardware lasting. You can move copy the VM image file to a DVD or hard drive and plug it into any Mac, Linux or Windows computer that has a VM player and boot the VM image and be up and running. So your system becomes abot as "portable" as a music CD.

Buy him a few 20" iMacs and set up his software in VMware's "Fusion" and he'll be good to go. He will be able to keep up to date as new version of the OS come out while is old software runs in the VM.

Question for the vet

[pharmaguru]

Please ask your Dad where I can find a dog that will live for the next 50-60 years. I mean, if he's expecting his new PC's to live five times longer than average, I should be able to find that in a pet, right? I know he probably doesn't like all that fuss changing machines and paying somebody to set everything up and retraining on new software every three or four years. What a bother! But, I don't appreciate paying somebody every 10 years to help set up a new dog and get him broken in. All those puppy shots, peeing on the floor, obedience training, missing socks and chewed up slippers. I figure if I can find a dog that'll last five decades I'll just stick with that. Oh, it'll run really slow towards the end of its life but it should still be reliable enough for me. Maybe he has some other tips... What parts would he expect to wear out first? Should I keep any spares on hand? Should I try to find one with no moving parts in the first place? Will vacuuming it out every year or so help?

Not on the net

[tepples]

mmarlett wrote:

(even if it runs off of one machine as a server without the Internet)

CarpetShark wrote:

Websites on the net require constant security upgrades

Am I missing something?

you're also doubling the number of systems to maintain by moving to a client/server architecture.

Not necessarily. I know of a few programs that run as a service, exposing a web server that listens only on some port of localhost. Freenet is one of them.

Re:Not on the net

[CarpetShark]

Am I missing something?

Yes, you missed this all-important qualifier:

even if

What is this crap!?

[dave562]

I'm completely convinced that the editors have gone off their rockers. I submitted an Ask Slashdot question a couple of weeks ago. I wanted to know about how to integrate Linux and Windows in an enterprise network. I figured that not only could I learn something, but so could a whole slew of other people. That question got rejected. I guess nobody cares about interoperability between (arguably) the two largest operating systems in the enterprise space. So my question got rejected, but this tool (nothing personal OP) wants to know how to replace a 486 and a Pentium with something a bit more modern? And that question makes it to the front page as a worthwhile story? Jesus fucking Christ people!

ideas

[ecloud]

If he wants to use the same DOS software I guess you could install FreeDos which might be less flaky than Windows95. But I suppose he uses Windows GUI features too, he doesn't just run the DOS programs from the command line?

Anyway hardware-wise (again if the software requirements are the same) maybe choose one of those Atom-based compact systems, hoping that low power requirements and low temperature will translate to longer life. Just make sure the board has all solid-state caps, no electrolytics. Boot from a solid-state drive or a CF card on a CF-to-IDE adapter. I'd be more inclined to build it rather than buy a prefab system, in case the power supply sucks and you need to replace it with a more reliable one, or something like that.

Somebody else suggested emulation, also a good idea but I doubt you need to spend money for VMWare. Maybe pick a Linux distro that doesn't suggest constant downloading of updates, so once you have installed it you won't expect it to break later; then at least he can run Firefox alongside the old DOS apps, on Bochs or dosemu I guess (has been a while since I tried it). Make sure you don't enable swap if you are using an SSD, maybe also pick a distro that puts frequently-written files (like /tmp and some stuff in /var) on a ramdisk. Put in enough memory so that the swap won't be missed.

new software

[Anonymous Coward]

For open source software there's evette.

http://evette.homeip.net/cgi-bin/makepage.py?menu=homecode.dat&targetpage=home.dat

If he was fine with this one...

[BoneFlower]

for that long, just about anything currently on the market should be fine.

Others are discussing hardware, I'll discuss the OS choice a bit.

A 64 bit OS might be worth considering, in the timespan you are dealing with they will probably become standard. He might be fine keeping the computer, but at some point he might need a software upgrade that requires it.

As much of a mac fan as I am, in 10-15 years, who the fuck knows what sort of processor they will be using. I'd be hesitant to recommend one for such a long service life due to their propensity to change processors more often than that. Scratch them off consideration entirely if they are even on your list.

Linux is a possibility, if all the software he needs(or equivalents) is available. This can work now.

Windows Vista... Some find good luck with it. If this needs to happen now and he needs or demands windows, research heavily the hardware and software that works well under Vista. Vista can be made to run well, but there are a lot of crap drivers and a lot of applications that really don't like the security features.

If you can delay the purchase, consider Windows 7. It's like Vista, only done right.

You could get any old thing and run XP, but looking that far into the future, you really should be looking at a 64 bit OS. Sure, any hardware you get now should run a 64 bit OS ok, but it would be best to plan from the start specifically for it.

Re:Vet Software

[mr_mischief]

Hopefully the DOS version was implemented by a software company that has since implemented newer versions and has provided their customers an upgrade path. It's hard to tell when buying software for a vertical market like veterinary office management which software house will still be in business 15 years after you buy their product.

This is one of the main reasons for Open Source Software, and particularly for the FSF's idea of Free Software. If someone buys a software package and has the source, another programmer can be hired to update or port that software. Without the application's source, we're talking about either supporting a legacy environment for the sake of the package he needs or finding a way to get the data out of what could be a convoluted proprietary binary storage format.

I'm not saying that all software needs to be open and have lots of liberal license terms protecting the buyer. I am saying that not having those extra license terms protecting the buyer carries a cost in some situations that has nothing to do with the upfront charges for the package. Those costs need to be weighed and considered just like any others.

Your fathers cheapness

[Jinjuku]

I know he's your dad, but he's cheap. What needs to happen here is 'expectation' management. The real issues are SUPPORT for that DOS based application. What happens when the data gets hosed and even the backup's are corrupted? You are doing him no favors. Set his expectation now while you still have a chance. Machines running for 15 years are simply by the grace of some non existent higher power.

dust

[Anonymous Coward]

I would like to second the mention of "DUST" as being one of the major factors in PC damage.

Overheating your system will greatly reduce or stop its operational ability.

On a modern computer the CPU will not last very long once the fans stop due to excessive dust.

Move away from Microsoft - Use Open Source

[Anonymous Coward]

Moving away from Microsoft Based technology would probably be your 1st step.

15 years out of todays computers?

[bored]

Give me a break, the quality of capacitors, and the push for ROHS (see http://www.google.com/search?hl=en&rls=com.microsoft%3Aen-us&q=tin+whiskers+ROHS&btnG=Search) among other things has definitely put an age limit on most modern hardware. I have machines from the 70's that still run and 486's from the mid 90's that run continuously (firewalls). But strangely enough, my more recently machines rarely last more than 5 years.

With the current price of machines, and their apparent disposablity, you probably better off trying to create an environment that isn't dependent on any piece of hardware lasting more than a couple of years. Spending thousands on "quality" hardware isn't any guarantee of lifespan. Your probably better off buying a year or two old model (buy it new, the idea being the bugs are worked out) of a low end business class machine from dell or HP, lenovo, etc.

Then virtualize your environment. That way you can just swap the a dead machine for a working one in the future without having to worry about compatibility.

System

[Anonymous Coward]

I would get: Intel Atom 330 CPU+chipset. High efficiency, fanless power supply. Small SSD (maybe two in software RAID, to protect against failure). Debian GNU/Linux, with a maximally minimalistic install for what he needs. Nice LCD panel. USB keyboard+mouse. Very low power, no moving parts. Should not be too expensive either.

Deponds..

[Junta]

In Linux, RAID5 is equally available in software as RAID1. The controller requirements are equal, no third party software required. This has been the case for many years. In Windows, I'm not sure what the situation is, as they at least restrict their workstation/desktop platforms to not have that feature built in and my experience in that realm is limited outside that scope.

My most recent build, for example, I did 4 drives of 750 GB drives for an approximate capacity of 2.25 TB (counting the manufacturer way). RAID1 by itself isn't going to do it, and 'RAID0+1' with 4 1 TB drives gets you to 2 TB of storage at increased cost. The cost/GB goes up from 750GB to 1TB currently, and you are paying for 4TB at that increased rate rather than 3.2 TB of storage at the cheaper rate. I really fail to understand the point of 0+1, it reduces the reliability compared to mirroring, isn't much better than RAID5 reliability-wise (if the *right* two drives in a 4 drive 0+1 setup fail, you will make it, but any other combo would fail just the same), and yields smaller capacity. Admittedly, the IO performance can be very good in that situation, but most applications do not extract enough benefit from that to make a difference for most uses.

If the desired capacity is smaller than the generally cheapest drives, 3 drives for RAID 5 is pointless. Anything other than that and you have to start doing the relatively trivial math to figure out the cost for mirror vs. raid5 of lower capacity drives to achieve the same end. For most people if 0+1 starts entering the equation, they would be better off with RAID5/6 with faster drives than 0+1 of cheaper slow drives.

Don't do Consumer PCs

[Lally Singh]

Consumer hardware just isn't built to last. Consider some used-but-in-good-condition workstations. E.g. an IBM POWER box, an UltraSparc box, or what was a good x86 server a few years ago (though I'm less confident on the x86 front) . The chips are built to last longer, as are the rest of the parts. Even the drives are selected out of the better lots from the manufacturers. A small RAID array (e.g. just 3 mirrored drives) may not hurt either. Considering how little data he likely needs to store, small, high quality drives wouldn't be terribly expensive, even if they're SAS.

When your customer base is paying for support, you cost-minimize downtime to make them happier and to reduce your own support costs.

Virtualize.

[peacefinder]

In addition to whatever else you are doing, seriously consider using virtualization to de-couple the information systems that actually Get Stuff Done for him from the underlying hardware. That way, when migration time comes around again, you can just slide the VM from one hypervisor+hardware platform to another. His systems turn into hermit crabs, changing hard shells at need without any painful and dangerous molting.

dumb terminals

[capsteve]

fact is that the components that fail the quickest will be anything that has any kind of mechanical movement, i.e. hard drives, fans, switchs, keyboards. i'm sure he's replaced a few fans, harddrives and keyboards in the last 15 years. why not upgrade him to a wyse terminal or other thin client, run terminal services, maintain the same software system he currently is using, and upgrade his two system network into a three system network, i.e. a server in a closet, and two wyse terminals. build/spec the server to include a backup hardware and software, raid 5 in a minimum 3 disk config, and a ups with shutdown software(powerchute).

Virtual?

[claytonjr]

If you want to make a setup last for 15 years, please consider going virtual. Something like VirtualBox would be good for DOS and Windows 95.

When the virtual server gets old, just move the virtual hard drive files over to a new server. This setup can last longer than 15 years.

Not cost effective.

[jonadab]

You can *try* to design systems for that kind of lifespan, but it's not cost effective. You end up with so much redundancy (redundant UPS, redundant power supply, redundant backplane, redundant hot-swappable hard drives with online spares, ...) that the cost is unnecessarily high. If high-availability were a major concern, because, say, an hour of downtime would cost you millions of dollars, then you would do it. But for a small veterinary practice, that's just not the case.

Set up a system for daily offsite backups and be happy. It's so much cheaper, that if your hardware fails every three years, you still save money. Hopefully it won't fail quite that often most of the time, so then you save even more.

late reply

[xalorous]

By all means build a computer for him.

Skimp on:
Case
peripherals
video (use onboard)
sound (use onboard)

Buy decent: better than average, but don't go cutting edge...costs too much
RAM - 3GB unless you use 64 bit OS and apps, otherise go 4 or 8
motherboard
processor

Spend extra on:
Hard Drives
Fans and Heatsinks
Power Supplies

And have a plan for when hard drive, fan, or power supply fails.

For processor, use a Core 2 Duo, 3GHz or higher. Reason? Simple, save a ton of cash over i7 or quad core. Pack it with RAM, but if you're not using a 64 bit OS, over 3Gigs is a waste.

As for software, sounds like he's running a database based system. Highly recommend looking over at one of the shareware repositories looking for software that does what you need it to do. He cannot possibly be the only veterinarian running a small business who needs "veterinary/medical record keeping + shipping & receiving + payroll" software. Someone has done this before.

Consider 3 or 4 machines. One for the database/file server. One for the front desk where the work is done. This one gets a nice monitor keyboard and mouse. The third is the backup machine. Server and backup machine need to be Raid 5 for redundancy, and the backup machine should be offsite. An online data repository might work for this, depending how much DATA needs to be backed up. Don't backup the OS and apps this way, a ghost type image taken after install+update is the best way to back those things up. This keeps the backups small.

Athlon XP is the best cpu evar

[sheaman]

just do linux on an AMD Athlon XP. they're hell of reliable processors, powerful enough, i've been running one for 7 years, overclocked, broken heatsinks, treated horribly, but it still holds up and has a 2 year uptime. i think virtualization sounds good, just run vmware or qemu inside of a STABLE linux distro (debian, ubuntu LTS, slackware)