Slashdot Mirror
How Often Do You Replace Your Hard Drives?
Telemachas asks: "I recently purchased a Dell P4 2.8 GHz swap meet computer with a 200 gig hard disk for a good price and all is working fine. It does not seem prudent, however, to trust my data on a swap meet item. For another @ $ 75.00 each I can purchase new 200 gig HDDs. I would also like to do my first RAID system. I am now wondering how often, if at all, do Slashdot readers replace their HDDs?"
When they break?
I'm still using a 40 Gig HDD that came with a HP system (not in the same system any more) for the last 5 years. It's a Seagate. But I've used other drives that I've simply disposed of due to limited size and space in the tower that lasted for even longer.
That which does not kill me only postpones the inevitable.
For home, I never replace a drive unless one goes down. I just have one drive backup to the other (and vice versa) at night, then store my important files at work.
At work, we have everything setup as Raid 1, and only replace drives when they go down, which is rarely. Not sure if this is the best approach, but considering we take offsite incremental backups every 15 minutes it's not really a catastrophic event even if both go down.
Crack - Free with every butt and set of boobs
I replace them when they die or I need more space.
Yay, I have a sig.
One day too late
Preemptively replacing hard drives is dumb, sorry. Back things up and replace them when they die, because they will.
I hardly ever replace a hard drive. I mean...sure. I've replaced a couple when I make housecalls or whatnot, but I have harddrives on a Gateway 2000 that's past the decade old line that works perfectly. Sure it had a gig and a half, but it works as an internet terminal that runs linux. I've upgraded hard drives, but always keep them around (the magnets are fun to play with) and I am always sure to have a back up or at least multi levels of redundancy to keep things safe.
Yesterday I replaced my 20GB Fujitsu drive that I've had since 2000. It's still working fine, it just got a bit small. Harddrives are actually astonishingly reliable.
At least every 3 months. By then the iron on the platters is already starting to oxidize. There are probably errors already being corrected for every 100th sector read, and likely some sectors have been remapped to a different physical location (rather than all in a neat contiguous length of sectors), which slows down my PostgreSQL sequence scans when the table is out of cache.
after like a year of constant 140F-ish temps, my laptop hard drive finally died. Then I replaced it lol. Other than that why would you put yourself through the hell of reinstalling XP and all your software? Imaging rarely works cuz of XP protections for OS drives and you'll often find yourself talking to Indian people about why you're reinstalling XP. There's just no reason to replace them until they break, just back up your data often.
Google's Super Secret Search Algorithm: SELECT @search_results FROM internet WHERE @search_results = 'good'
for serial. I keep any pertinent personal data on other local drives than the OS itself, and then replicate that data out to my file/backup server, which does raid. If at any point any drive in there fails, I'll replace it, but not until it dies. Drives shelf lives are variable, and I want the most for my buck.
How Jaded Are You?
never
You should really be careful with those magnets. They're extremely powerful.
When I was in college, I knew one guy who took them out of a drive, and was playing with them. This would have been back around 1992, when drives had to use far more powerful magnets than they do today. In short, he somehow got them too close to his scrotum. The two magnets snapped together, crushing one of his testicles. The doctors couldn't do anything but remove what was left. Of course, we laughed at his misfortune until we graduated and moved on.
(hi, using AC since im too lazy to log in)
I found the Seagate 300 gb 16 meg cache IDE drive to be at a nice price point ($95 or so), so I popped one of those in my old crapbox. It runs pretty damn nice, so I added a 400 when those were on sale with crappy rebate (about $110, iirc).
Seagate is good about their rebates, actually, I have gotten them quickly.
The 5 year warranty is not bad, either. Just do as the company says to and use a cooling fan on the drive, they love to run hot. I have yet to have one fail on me, but since drives are moving parts there is always the possibility.
I would not go back to smaller sized drives, it seems like once I add a new one it fills up fast.
Sure, sometimes an old drive can be trusted, but only as far as it can be thrown. What old pulls I have played with, I've wasted some major brands (maxtor, wd, maybe a quantum too) but I have yet to waste a Seagate....YMMV~
Outpost.com usually has some nice deals on retail boxes.
I started building computers twelve years ago.
:)
The only drive I've had die before I retired it myself from sheer obsolescence was an IBM 20GB "DeskStar" model; this happened about five years ago, IIRC. The drive made noise and froze the system when I would read particular files; to my frustration, it occurred when I read some of the files that were important to me (documents, programming projects, one folder of MP3s, etc.)
My solution was to put the drive in the freezer for a few hours; UNBELIEVABLY, it worked - I would have about ten minutes to copy as much as I could off the drive before it would start making noise again. I got most of what I needed off of it.
Incidentally, IBM was very good about the whole thing; they sent me a new drive the day I called them. Too bad they sold their HD division to Hitachi...
Anyway, I've had FAR worse luck with power supplies; I usually go through one of those every other year. Recently, ALL of the drives in my RAID 5 array (4x 120GB Seagate drives) as well as a fifth one (an identical Seagate 120GB that's standalone) started making noise at around the same time; of course I assumed there was some defect with this particular drive model.
But thankfully, it turned out only to be my power supply (the +5V line would deliver +4.4V ~ +4.6V, while the +12V line would fluctuate between +11V and +13V). I can only conclude that Seagate drives are less tolerant than IBM/Hitachi's of power supply fluctuations, since I also have an old 80GB IBM/Hitachi Deskstar and a much newer 250GB SATA IBM/Hitachi drive, and neither batted an eye.
Likewise, the system showed no other symptoms that pointed at the power supply; so a week or so ago, this post would have looked very different, with a few "F-You Seagate"'s thrown in there.
I vote for this as one of the dumbest questions ever, replace them when they die, before is a waste of money.
Sorry, teleporters just kill you and then make a copy. A perfect, soul-less copy.
When they start to play the violin, it's time to kick their butt outta the case.
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
I replace my hard drive when the S.M.A.R.T. info starts to signify problems, such as too many relocated sectors.
Hard drives are the single most common point of failure for hard drives.
... I just wish people were better about following through with RMAs. They just take so darn long.
You really see it in laptops. Remember how quiet your laptop was when you first bought it and started it up? And now, a year or two later, you can hear it read the hard drive from across the room? Keep using it and you'll get these surprisingly loud clicks and soon things will be a-corruptin'.
Now Linux nerds with closets full of beloved "vintage" computers might find their hard drives last forever, but those of us who use our computers find that only rarely do hard drives pass their third birthday without some bad sectors spreading like cancer.
They're cheap enough that business travelers I know are putting their important data on flash drives and treating their laptops' hard drives as essentially disposable. Road warriors are lucky to get a reliable year. Fortunately a new laptop hard drive is less than a new laptop battery.
As they miniaturize they become more fragile
As I've already seen a couple of people say, don't preemptively replace your hard drives.
Allow me to add: Here's why.
Hardware failure rates follow a curve on average. They fail a lot after initial purchase, then slope down to their minimum after a couple of [relevant time periods] (probably "weeks" or "months" for hard drives, varies by what kind of thing it is), then slowly slopes upwards again.
(Please do not miss the phrase "on average". Certain specific flaws can cause a certain product line to have unusual characteristics, like a sudden spike at six months or something. However, unless you somehow figure out a way to guess which hard drives are going to have such failures in six months when it's pretty amazing for the exact same hard drive to even be on the market for six months, the fact that these things can theoretically happen can't have much impact on your decisions. After all, if you knew that was going to happen, you'd just plain not buy the drive, period, regardless of the argument in this post.)
Therefore, if you've got a "burned in" drive, you will be replacing a known-high-reliablility component with a component with a lower expected reliability. (I use "expected" in the probability/statistics sense here.) Unless you've discovered that you do have one of those funky products that all die in ten months, this is a bad move on average.
I replace hard drives when they fail. I try to act as if they could die at any minute, although I fail.
(But I try to get better. I'm in an all-laptop house, so it's difficult to have the convenience of an integrated backup solution and an automated, unforgettable script. However, with the recent Linux kernels finally supporting my SD card reader, I've gotten a high-capacity, slow, cheap SD card to stick in the previously-useless slot and I have an rsync now backing up the files I'd cry if I lost every hour. Sure, 1GB can't backup my entire system but most people's "cry if I lost it" datasets would fit into that. (Yes, there are exceptions... but if you're one of them, you've already got another back up solution in place, right? Right?))
Running Knoppix on a dumb terminal with only a cd-rom drive, network card, motherboard, etc. without a harddrive, and then backing up everything onto a server over a broadband internet connection. Off site data center takes care of data backup, redundancy, etc. No mess!
First year in college, 2 maxtor 250's died. RMA'd and both warranty replacements died. Then the HD in my laptop died. Then one other random HD died. That's why I have everything important on at least 3 hard drives and use RAID5 for all of my general storage.
I don't upgrade single drives at a time. I have dedicated file servers to put the majority of my data on. The first was 8x20GB drives, then 8x120GB drives and my current is 8x250GB drives. I rebuild when I run out of space and can afford the upgrade. When I do, I take down the old system and have several drives to throw around in spare systems and friends computers. This happens every few years I guess. The file servers are all RAID 5 and I upgraded to a gigabit network with the last one so it's pretty speedy and redundant. It's also handy when you have data to share between several computers and several users. Though, I believe my next system will simply be a MacPro with 3x750GB drives. I'm getting to the point where I wish the majority of my data was on my computer locally so I don't have to worry about permissions and resource forks. I'm also getting tired of the whole second-computer-for-data thing. I'm ready to consolidate. I guess I'll finally have to do decent backups though in case a drive goes down.
The last time I had to replace a drive (in my own system) was 3 years ago. Not sure what happened, just came home one day and my computer wouldn't even post with that drive installed. Was a 60 gig Maxtor that was still under warranty, they replaced it with a newer model. I still use that drive to this date along with a 120 gig thats about 2 years old and a 300 gig sata drive I got last month.
From what I've seen harddrives have a very good life expectancy for electronics with moving parts. I know people still using 8 gig drives that don't show any signs of failure. I even have a very old laptop with a 800meg drive that still boots and doesn't have bad sectors.
What I do in cases of a used drives is check for signs of failure, bad sectors, file system corruption ect. As well as cycling the system to a full power down ware the drive stops spinning a good 5-6 times in a row. (being the local computer hobbiest I get asked to fix or build computers for people that don't have the money for new)
And recover them from the backup. You do make backups don't you?
POKE 36879,8
I've got drives older than I am sitting around that still work. Granted I don't use them anymore, but they ran for quite some time.
;)
As most people have said, the best thing to do is backup a lot, and replace it when it starts to go bad. If it starts making loud crazy noises, chances are its on its last leg. If you get random boot errors...same thing. Basically, when it starts fucking up, its probably time for a replacement, but beyond that...well...just ride it out. Don't spend money you don't need to spend just to keep on a schedule. As long as you back up things that are terribly important you should not have a problem.
For the 80 gigs of pr0n well...its not like you watch ALL of it anymore anyway right? You're gonna get new porn regardless so its probably not a huge loss if that all goes down with your HDD. Then again, they are coming out with some rather large new removeable formats (Blu-ray or HDDVD) so maybe its not as hard to backup as you'd think.
"Those who would sacrifice essential liberties for a little temporary safety deserve neither liberty nor safety." - BenF
The redundancy buys you reduced downtime in the event of most failures. Go with multiple RAIDs in different systems (or cities!) for backup.
Until they start sounding funny, generally, but I always make backups of real data.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
I've never had a hard disk die on my personal computers (Although I seen dozens of dead SCSI drives in servers) and never needed to replace one. If I start running low on space I just offload to firewire drives, DVDs, etc., but mechanical failure has never been an issue.
I've only had two hard drives die on me. The first being a 12 GB hard drive that died 8 months ago that was over 7 years old. The second one, I fried when I hacked together a machine for my workbench. What may be your best option would be to get one backup drive, internal or external, and use it strictly for backing up your important data once a week or so. This way you have a backup hard drive that is rarely used, and you can replace your drives if they die.
I third that
Never start replacing components unless it's the power supply or fans. Normally once my hardware starts screwing up I just sell the whole thing at a swapmeet as generally all the components will start all screwing up together.
Err, good luck with your new machine.
Australian running a company that does C# / C++ / Java / SQL / Python / Mathematica
I have about 3 100gb drives, 1 300gb drive, and 1 200gb drive.
I recently dumped them all in favor of 2 fast and quiet 320 gb sata seagate drives. First, because of power and heat concerns (I figure 2 hard drives compared to 5 would have less heat concerns) and second because of noise (the maxtor and samsung drives seemed unusually loud, like they were about to fail).
For some reason my crappy sony DVD recorder didn't like sharing an IDE with a hard drive (yes, I tried all the different master slave and cable select options), and my new motherboard only has one IDE slot (I have no idea why), so I had to replace all my IDE drives with sata drives anyways.
I figure these will last me a good 5-6 years unless WD releases a 10,000rpm 3gb/s raptor drive that is both quiet and blows these drives out of the water (the raptors are faster, but not by much).
However, my main point is that if you are upgrading to Vista (yah, I know, this is slashdot), don't upgrade now if you are patient (I'm not). Samsung will be releasing a flash hybrid harddrive compatible with vista (and perhaps linux) in January that should speed up boot up and cache times.
When the drive breaks, starts to run slow, or has other problems, I replace it. When it runs out of space, I get an additional drive (and throw out the oldest/smallest/junkiest one I have).
Never underestimate the power of simplicity.
This post expresses my opinion, not that of my employer. And yes, IAAL.
I replace them only when they fail or if I need more space. Seriously, hard drives are getting cheaper every day. Why buy ahead if you don't need to? My home server's system drive is a 13GB Maxtor that I bought in 1998. I have Debian and swap installed on it. I keep all of my data on six 200GB drives with software RAID5. Sure, the 13 gig could die at any moment so I keep backups and run smartmontools to help warn me if it's about to die. But if it's not broke, don't fix it.
Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.
I dont replace them, i just relocated them, to my wall of shame (DeathStars {ibm}, Maxtor {extra crunchy}, Seagate {only my controller failed, go fig}) .....one torx screwdriver away from fingerprints
Check out smartmontools (http://smartmontools.sourceforge.net/) it is a very good way to keep track of drive health... but backups are always good ;-)
From the time of my 20MB HD in my first 286 until about the time 40GB drives were common, I didn't have a single drive die. I think maybe a couple sectors on one 2GB disk, but it never got any worse. I used to laugh at people who ever had a HD die.
Well ever since disks have gone over 40GB, I've had nothing but bad luck. I'd guess across all my systems I have about 8 HDs running and have a HD die on me about every 6 months. I try to ensure that all the drives are well cooled and that the case is solidly mounted to elimiate vibrations, but they still die. HD prices have dropped, their capacity increased, but their reliabilty dropped. Now I just buy drives in pairs because I can't trust drives anymore.
On top of everything being RAID to make up for crappy HDs, I still had storage issues but this time getting burned by a bad SATA controller. It occasionally wrote corrupted blocks to one of the disks in the set without erroring, but it was definately a noticeable problem when files would randomly be corrupted.
Best way to sum it up: don't trust your HDs to last, don't trust your controller, don't trust your RAID, and don't trust your backups. Be as paranoid as possible with your backups by making sure you can actually read your own backups and make more then one backup.
REPLACE hard drives? I usually just keep them around until I get tired of the knocking noise, after which I use the magnets to build GSM signal jammers, because them damn cellular phones keep putting a weird clicking noise on my sound system. Then people are all whiny cause their phone doesn't work, and I just don't know why :) I even suggested to someone that they wrapped the arm the held the phone up in with tin foil and they actually did!
I've been using home computers since before they had harddrives, and in all this time I've had only two failures. ... I just did a quick count of current and retired drives in the house: 17. That doesn't include drives I've put in gift machines, or tossed in periodic clean-outs. (Overdue for another, yes.)
Perhaps I've been lucky, but I've also come to suspect that people with chronic failures have bad power supplies and/or tight, hot cases. Keep your drives spaced and cool.
So I just buy new drives as the size-price ratio makes it absurd to not upgrade, and retire the old ones farther down the ide and into secondary machines. Eventually a drive is too small to bother with, like, these days, anything smaller than a DVDRW.
That normal upgrade cycle has kept me and friends and family ahead of any old-drive-failure point there might be.
my, oh my. I never thought someone would take the phrase "water-cooling" a computer to the extreme, and actually submerging the whole damn thing in water. I am blown away that your harddrive actually rusted to death, and didn't stop working immediately. Do you have a connection with a HD manufacturer's secret R&D department for waterproof harddrives that you aren't telling us about?
I have had to replace a few hard drives (in multiple machines) over the past few years. Maybe 1/4 of all my machines have had some hard drive problem. I've come to the point where I always have a regular backup system for all of my machines. If I could have a RAID-1 setup on my laptop I would. I have software RAID on most of my desktops/pseudo-servers. Maybe that's excessive (if it is non-essential stuff then a usb backup is enough) but I've had enough bad luck that I'm sick of losing data. And hard drives are just so cheap nowadays.
What comes first, finding a teacher or becoming a student?
It's been my experience that hard drives will either fail within a year or two, or will last quite a long time.
Although I know from an engineering standpoint, that old drives would appear to be more prone to failure, I've observed that drive failures tend to be randomly distributed events. In other words, a new drive may be just as likely to fail as an old drive -- it's just a matter of odds and time.
Swapping out an old drive for a new one does not even necessarily reduce the risk of failure. Many drives fail in their first months of operation. QA certainly isn't what it used to be.
If you're this concerned about the integrity of your data, you should be making frequent backups and/or using a mirrored RAID setup. With RAID you don't need to worry about drives aging and swapping them in and out, because you lose nothing when a drive fails, allowing you to wait until it does so, eliminating any guesswork. When a drive fails, you replace it, and in the meantime the spare picks up the slack. As long as you've got a clean power source, and replace the dead drive in a timely manner, the odds of the other drive failing in the interim are miniscule. (If you're *really* worried, you can jump up to a 3 or 4 drive configuration, although at that point, you should probably be considering some sort of tape backup solution instead to cut out the power supply as a variable)
In other words, don't sweat it. Keep an eye on the SMART data, and use RAID, and for all intents and purposes, you can sleep soundly at night.
-- If you try to fail and succeed, which have you done? - Uli's moose
Rule number one: always keep an extra drive around. Drives are cheap, and they die regularly. Also, the cost of buying that _one_ extra drive is constant. You always have an extra drive around. It's not like you have to buy two each time you go to the store. You drives will die at 8pm on a Sunday night, just before you go on that 3-week business trip, otherwise. I promise.
Rule number two: never spend more than $100. The best $/GB always seems to me to be in the $100 range these days. I usually make sure to pick up drives at Fry's whenever I see something substantially larger than what I have now for less than $100.
Rule number three: Stay ahead of drive failures. If you have important data on those crappy, cheap $100 IDE drives, replace them every two years at least. In those two years, you can double your capacity for less cost. Use the old drives for backups of important stuff, just in case a newer drive bites the dust. Or, leave it as-is, and use it like a snapshot of your working data.
God... I have a 10MB hard drive (it's part of an IBM XT, which I got from my maternal grandfather) that, to the best of my knowledge, still works. Though that whole system's been sitting out in the garage for a couple of years now, so I don't know how reliable it is anymore. Generally I don't worry about replacing the hard drives in a system - when a system becomes too old or otherwise nonfunctional, every functional hard drive (and even marginal ones, if I have a Linux boot disk - dd is great) gets migrated over to the new system if possible. This is why I've had to actually buy an ATA controller card - the current system has three HDs and two DVD+/-RW drives in it, all ATA. (It's also about five years old - older Athlon OCed to be about a 2200+, 1.25GB RAM, Win2K, etc.) I'm so not looking forward to the next new system I get - it'll eventually be dealing with something on the order of 8-10 SATA drives, I imagine...
As someone who uses drives 24/7 (they are on all the time) it is better to sell them on ebay before their warranty is up. Now that Seagate has 5 year warranty on their standard drives, always sell them at around the 1.5-2 year mark and upgrade to newer drives. If you don't use your drives 24/7 and only use them much more rarely then you can get away with not replacing them longer but...
I have learned the hardway from when I was younger: If you can afford RAID, even simply mirroring, do it. I use RAID 5 and there is no way I'm ever going back to a non-RAID setup, you save loads of time in not having to back stuff up to CD or DVD. Although you should ALWAYS back up smaller important files to either 1) Flash or 2) CD/DVD.
Some information is irreplacable, never get complacent. Especially since computers are now the nerve centers of our lives in many respects, they hold countless hours of work and irreplacable memories (photo's, etc).
This may go against the grain here, but I replace my desktop drive about every 12-18 months. As I see it, here are the benefits of doing so: +1) The drive still has decent resale value at that point, particularly if you sell on a computer forum and not on ebay. This helps reduce the cost of the hard drive update. +2) Drive capacities are increasing quickly while costs continue to decline. This reduces the cost of the upgrade. +3) Replacing before the warranty period is up means that the likelihood of experiencing a hard drive failure is low. +4) While WinXP is a lot better than Win9x, it still doesn't hurt to do a fresh reinstall every 12-18 months. A hard disk replacement is the perfect timing for this. Of course, there are some valid counter-arguments to these points: -1) Security. (i.e., somebody could recover your private data.) I run Darik's Boot 'n' Nuke a few times, so I'm not terribly concerned about this. After running such a program, the odds of somebody successfully recovering data on a home budget are pretty low. -2) You may be replacing too often. Well, I can't do much about that. But good drives don't cost much more than $100-$150 these days. A little peace of mind is worth something, and the regular size/speed upgrades are a nice bonus. -3) This is no substitute for backups. I completely agree, and make backups of my most critical data to remote servers. -4) Perhaps this isn't necessary. Perhaps not, but a fresh format is a helpful after 18 months. Any way around it, I acknowledge that this strategy is a bit more expensive than may be necessary, but it has served me well in the past six years +. I've only had one drive fail in the past, back when I let my drives go well beyond the warranty period. Of course, that drive was a total loss, with no recovery of value to apply to the new drive, and there were some non-recoverable files. In my opinion, preventing problems before they occur is preferable, and getting speed and capacity boosts are just icing on the cake. -- Paul
OpenSource.MathCancer.org: open source comp bio
PC Power & Cooling. Don't bother with anything else. I have yet to have one die on me, and I've been using them for around ten years. Their Silencer models are awesome.
because honestly, what's the point of defragmenting a hard drive? It's just going to fragment again. And don't get me started about journalized file systems, that'd make too much bloody sense!
I accidentally posted my comment (meant to click preview) without inserting my formatting tags. Please disregard and read this instead. Sorry!! -- Paul
This may go against the grain here, but I replace my desktop drive about every 12-18 months. As I see it, here are the benefits of doing so:
+1) The drive still has decent resale value at that point, particularly if you sell on a computer forum and not on ebay. This helps reduce the cost of the hard drive update.
+2) Drive capacities are increasing quickly while costs continue to decline. This reduces the cost of the upgrade.
+3) Replacing before the warranty period is up means that the likelihood of experiencing a hard drive failure is low.
+4) While WinXP is a lot better than Win9x, it still doesn't hurt to do a fresh reinstall every 12-18 months. A hard disk replacement is the perfect timing for this.
Of course, there are some valid counter-arguments to these points:
-1) Security. (i.e., somebody could recover your private data.) I run Darik's Boot 'n' Nuke a few times, so I'm not terribly concerned about this. After running such a program, the odds of somebody successfully recovering data on a home budget are pretty low.
-2) You may be replacing too often. Well, I can't do much about that. But good drives don't cost much more than $100-$150 these days. A little peace of mind is worth something, and the regular size/speed upgrades are a nice bonus.
-3) This is no substitute for backups. I completely agree, and make backups of my most critical data to remote servers.
-4) Perhaps this isn't necessary. Perhaps not, but a fresh format is a helpful after 18 months.
Any way around it, I acknowledge that this strategy is a bit more expensive than may be necessary, but it has served me well in the past six years +. I've only had one drive fail in the past, back when I let my drives go well beyond the warranty period. Of course, that drive was a total loss, with no recovery of value to apply to the new drive, and there were some non-recoverable files. In my opinion, preventing problems before they occur is preferable, and getting speed and capacity boosts are just icing on the cake. -- Paul
OpenSource.MathCancer.org: open source comp bio
Chances are your motherboard will have a built-in raid chip. You'd be well advised to skip that chip and use a PCI raid card or some other solution.
Since the onboard raid chips are tied to a specific CPU family, you won't be able to move your hdd's to a new motherboard in a couple of years time. I just found out this week that the raid i'm running from an intel ICH6R chipset on an ASUS motherboard cannot be migrated to an Intel Core 2 Duo because the ICH6R chip doesn't support that CPU...
I remember, a young(er) me waking up bright and early on a saturday morning to begin preparations for a LAN party with the guys. Pepsi, hot pockets, plenty of extension cords, gamecu--... what's that clicking noise? !@#$%^&*()!!!
My first two hard drives were a 17GB Seagate and a 80GB Western Digital. Still using both of them (the Seagate was pretty cheap, but the WD was one of the first 7200rpm drives here).
Every drive failure I've experienced has had two things in common.
1. There were obvious warning signs(strange noises, etc) while the drive was still functioning properly.
2. The failure isn't sudden death of the drive, the drive has a fairly long period of almost working right.
Replacing the drive as soon as it starts to exhibit problems is much more important than worrying about the age of the drive.
When they fail... With Maxtors, generally that's warranty period + 1 week. Seagates and Western Digitals, generally warranty + 1 year. Of the 3 brands, Maxtor has the worst chance of not making it to warranty - I don't use them anymore.
In 03 I ended up getting a 80gig for Mothers Day, in April 04 I traded that for 2 raid'd ones before I moved (to another state), Mothers Day 05 I got a 120/160/180 I dont remember (its late Im tired and have a newborn lol), so now I have two in my machine and I still have plenty of room. :)PS Good Mothers Day gifts for a geek ;)
You only need one overwrite pass to obliterate data forever. Even governments can't recover it (except perhaps by means of non-computerised techniques). Fred in the Shed has no chance.
Given the way prices for each component of a computer system have changed at different rates with respect to one another over the years, if four-dimensional storage was at all possible, it must have been economically viable at some stage (bear in mind that until the advent of solid-state RAM in the mid-1970s, computers used all-magnetic storage), and there'd be a computer in a museum somewhere based on the principle.
Je fume. Tu fumes. Nous fûmes!
Saves a lot of hassle.
They BREAK!?! Holy shit! How am I going to protect my...uh..., vacation pictures?
Seiously, are there people out there who replace their hard drives just because they are not the new shiny model in the market? Always make regular backups of important data and that should be sufficient for most of the mortals - may not be applicable if you are running some sort of datacenter or some other important business where the minor efficiency (seek time etc.) fluctuation factor is very important.
I have much smaller hd config (40*1+120*1+200*2). No RAID or SATA, all old fashioned IDE cabled. 40 gigs one is my system drive and is about 5 years old (yeah I know thats incredibly inefficient for a system drive but it has served me well and I am not rich enough to have extra cash to go for a small smarter drive), all the rest of the drives are also about 2 years old, bought withing 3-4 months of each other. Never had any problems with any of them so far (and I use the 200 gb drives+USB casing as portable storage when I travel!) Just make sure to backup your important data. Blank DVDs are dirt cheap for crying out loud!.
Politicians and Pedophiles: Two groups of exploitive bastards who are most dangerous when they're thinking of children.
I have discs that are 13 years old and which are still running. Only that one and one from 1995 have any issues whatsoever - stiction in both cases.
When old ones broke or when I come up with urgent need of more filespace.
-Seeing the problem is ½ of solution-
It is so very simple. You replace your hard drive when a new drive under $200 can contain your last five OS installations plus data, er. might have to buy a secondary sub-$200 part to help a little. Gee thats five years for less than $400.
YMMV
I have two drives installed on my PC (w/ Windows) one is the applications drive and the other the data drive (though the biggest usually has two Linux partitions on it) and usually replace the smallest whenever it starts to get full often AND the price sweetspot for HDDs (i.e. the drives for which the price per Giga is lowest) is at a capacity two or more times bigger.
Since i mostly use my PC for gaming (hence using Windows), the size of games usually dictates the amount of space used in both drives (since the game is installed in the apps drive and i rip the CD/DVD image to the data drive to mount as a virtual CD/DVD).
I've only bought a desktop PC once (a long time ago - 386DX 20) and have been upgrading it myself ever since (only thing left from the original is the keyboard - with no windows key!!!!), so i've gone through the HDD (and other major components) upgrade process often.
In my experience, since i buy my HDDs at the sweet spot (thus never the biggest in the market) i get a new HDD roughly about every 1,5 years.
I've recently discovered external HDDs as a great cost-and-time-effective means of external data storage but i'm not including those in my calculation.
I have a 60GB drive I purchased four years ago still going strong. I have had a 40GB from five years ago that started clicking earlier this year. I have an 80GB drive that just won't cooperate for very long before it starts acting up. I now have a 200GB drive as a backup drive, it is only 6 months old.
In the system I gave to my parents it is still running a 20GB drive from six years ago with no problems. I have had friends give me old computers and the little 10GB drives in them are still working fine also and those drives are from '97.
Stay with what works. Backup often, and have a spare on hand for emergency replacement. Run RAID if you want to, (I plan on it when I get some spare money), and go with the percentages.
I don't replace drives until they break. What I do do however, is regularly run out of space.
What happens then? I buy a new large disc (last one was 300gb last week, 300gb is not the largest but at this time one of the best gb/size and big). I add this one to my PC, and move everything "up one disc" in the newer harddrives chain. Ok, that takes a night, but hey, let the PC do the work while I sleep. When my system maximum of 8 HD's is reached, I wipe the oldest one, remove it, and throw or give it away. When this is done, I usually have the few hundreds gig left accross drives to backup important data, and still enough space to download all these new TV shows (and pr0n ofcourse). Since I watch a lot of TV shows over the internet this takes a lot of space, and I don't put them on DVDs 'til the season is complete. And I have to burn them all twice too, because my sister is a researcher in Italy, and apparently, TV sucks over there. The burn for myself is usually a few months after that burn batch too. The most important data, apart from being backup up on all those discs is also regularly backup up to a removable drive which I then take to work to get backup up (hey it's MY company, so no issues there).
Ok, I may be a little bit crazy... but... nah wait no buts.
As they get full. Just kidding:-) But seriously, to answer the OP's question, I personally wil keep a hard drive for around 2-3 years depending on the brand. When I stop using a hard drive, I put it in an enclosure and sell it on ebay, or if it is a big drive I use it for a gaming console (ps2 + HDAdvance + flip top lid = good times) or to give to friends as gifts. Still, like it has been mentioned, there is no set time really...a hard drive can last 10 years, or 10 days.
Living With a Nerd
I was talking to one of the head technical managers for one of the US's climate data centers -- geeking out about setting up gentoo for atmospheric modeling, 10 to 100 TB datasets, etc. What he is now finding that they are loosing a disk or two about every time that they have to take sections of the cluster down for maint. Their numbers are looking something like 2 to 3 years for MTBF, but when you have hundreds of disks in a dataserver you can expect one to pop on any given day. This needs to be taken into account when budgeting for routine maintance.
For personal use, think about this... Computers are still roughly doubling their throughput/speed every ~1.5 years. Likewise diskdrives are increasing size and throughput. If you are upgrading your home box after 3 to 5 years, go ahead and upgrade the disk as well -- it is an electro-mechanical device that has already lasted past its MTBF. The are not that expensive when comparied to the data on the disk, and the time and frustration of dealing with a hard disk crash. WHat I do when upgrading is to retire the old disk (basically turing it into a backup)...
One of the other posters brought up a good point -- Infant mortality. Disk life expectancy is saddle shaped and most failurs happen almost immediately after plugging in (damaged in shipment). After that parts eventually ware out and you start seeing them die from old age. So, basically if the disk works for 5 minuites it is likely to run for a week. If it runs for a week it is likely to last a half year. It if lasts that long then there is a very good chance that it will last 3-5 years, and if it lasts that long it is time to throw it a retirement party for its dedication and service to the cause.
from my wife "there's no more room on the server for my mp3s..."
this is just a placeholder till i send back my real sig from the future.
I replaced the hard drives in my server at home (two drives, mirrored), simply because I was running out of space.
When it came time to build a small "jukebox" system to play music on, I decided to use one of the drives that had been in my server to save a little money. About a month after building the box, I started to get hdd errors. Ran fsck several times, but more and more bad clusters would appear. Eventually, I had to replace the drive with the other mirrored drive that had been in the server.
Same deal. Got about a month's use out of it, and then had the same problems. I then bought a new drive.
I don't know if it was a coincidence that I was able to replace the drives in my server before they both died at the same time, or if it was just the process of lying dormant for about six months before being put in a new box that caused them to die. The drives were about four years old.
I'm thinking now that even though I've got a mirrored system on the server (hence, I don't do backups), I better replace the drives at least once every four years, even if I haven't run out of space on them.
And believe me this is stupid. Get a HDD with a detachable interface, I had a HDD die on me once after a big power surge (Crappy power bar was upgraded soon after) and my HDD wouldn't boot at all, no BIOS, nothing. luckily the chipset on it (i'm not hardware keen enough to know exactly what that interface is called) unscrewed from the HDD itself....borrowed a similar model from work, put it on, put it back in PC and it worked, copied the files off it, got new better drive and put the interface back on the original HDD. Now it's something I consider when buying drives, are there interchangeable parts on it in case something dies.
Of course, real advice is to backup often. Which I also do, but I'm not a nightly backup kind of guy, more like every few months. That and RAID, cause it kills bugs dead.
When they die. And that's not very often. In fact, I've only ever had 2 drives die on me, and they were both Maxtors. I have an 80gb WD that I still use (for MythTV files, nothing too important) that is 6 years old. The Maxtors lasted about 6 months each.
Now I've got a 120gb Seagate, and 2 80GB WDs. Works nicely, will replace when they break... or I need more space, which may be soon... very soon.
I just backup important documents and my music is all on CDs if I should ever need it again.
Starmen.net
It all depends on the importance of the data you have on the drive. A hard drive is a "wear and tear" part and simply will not last forever. S'not a question of if...but a question of when. That being said I have an old Dell Poweredge 2200 that I think was bought in 1996 that's still running with the original HDD. I also have bought many DOA drives and had several die in a matter of days. It's a crap shoot. If your data is important and you can't lose it then invest in at least a RAID 1. If you're a geek and want to have some fun (and burn some bucks) then go RAID 5 or RAID 10. Many motherboards nowadays have some kind of integrated SATA RAID controller. If not then an IDE RAID card (PCI) will suit you for storage only...you won't get lots of speed out of it. Good luck.
...It's not 'incredibly stupid'. A question is a question. Just answer it if you have an answer, otherwise, shut up.
They're better when they're fresh! Just ask your local IT guy, he might even have some fresh drives which will be better than your tired old one.
Deleted
My drives don't die on a set schedule, so I replace them in more of a random timeline. I had a bout of bad luck a few years ago when they died only a couple months after purchase and I bought a number of them for my MythTV box. But after I stopped buying Western Digital it seems to be taking a lot longer to get to a death/replacement cycle.
You must be one of those Americans, I can tell. Buy something, use it, throw it away and buy another. Bah!
Take that hard drive down to your basement and open her up. Blow out all the dust bunnies (use your breath, not that expensive canned air stuff). File the head surfaces so they are new and shiny. Buff the surfaces of the drive platters. Replace the air filter with cotton from a Q-tip. Clean and oil (not too much now) the bearings. Resolder any spots that look like they have gone cold. Put it all back together and seal any open areas with a small piece of duct tape. Voila!
In my country, we don't just throw away a used hard drive. You sissy Americans.
I copy difficult to replace but not earth shattering files (mp3, video, etc) to another PC in another room over LAN every once in awile. Hard drive space is cheap, LAN is fast enough to do it overnight. It's unlikely both hard drives or both PCs will crash simultaneously. Of course they can both be stolen, or burned, but unlikely.
Really important data (pictures, financial data, etc) are backed up to dvd occasionally and stored elsewhere in the house, or the bank safety deposit box if you're really paranoid.
... is to, generally, not replace them until I see a data loss. (I'm pretty good about backing everything up; must be a habit I picked up at work.) I've had excellent luck with disks, though. After having a stiction problem with a 200MB Maxtor back in the early '90s, I've had only two disk failures since then: an old Seagate Hawk 2GB and a Compaq 18GB. I still have a couple of the 2GB disks running (in the firewall; they're big enough for that) and a slew of Barracudas. Those disks are at least ten years old. I'm comfortable running those old disks since I know their history: they were pulled from systems in perfect working order and bound for the dumpster at various employers who needed more space on systems. I'd be a lot more leary of using a disk that was in a system bought at a garage sale. There's no telling how many times the system/disk had been dropped and potentially damaged.
Now, having said that, I do have a project at home to begin replacing some of the older drives with higher capacity SATA disks. I'll need fewer disks and the newer disks seem much quieter. (A shelf full of those old Hawks and Barracudas have a distinctive "whine" that gets annoying to listen to for any length of time.)
CUR ALLOC 20195.....5804M
I think they used to make good PSUs, but they went astray somewhere...I've had two Antecs die on me in the past 6 months now (the last one with the same undervolt and pissed off RAID controller symptoms that you mentioned).
I decided to go with PC Power and Cooling 510W (~$200) for both my important machines. They have a (well-deserved, from what I know) reputation as having the best PSUs in the business. I would not, however, recommend going with the Silencer series as someone else in this thread recommended. They are basically repackaged Silverstones (which are themselves decent PSUs, admittedly) -- and do not meet the same specs that the PCPC "Turbo-Cool" series does. FWIW, you may see some reviews talking about excessive noise from the Turbo-Cools, but I think that was a previous revision - both units I got are very quiet.
...when I need a bigger one.
I can't quote a definitive source right now any better than you did, but I believe that the commonly accepted figure for drives (based on a lot of their components, such as electrolytic capacitors) is 50% per *10* degrees C. Still significant, but nothing to cause you to spend big bucks to cool it 2 degrees.
Leaving my pc powered 24/7, I typically start seeing disk errors pop up after around 3 years. I replace it when it a.) fails or b.) generates SMART errors.
>If you replace them on a schedule, you're still not guaranteed 100% reliability because a drive can fail way before MTBF...
It is a common misperception that MTBF ratings mean anything about how long an individual device is supposed to last. It's only a measure across a large number of units in total power-on hours, and only within the expected "useful life."
For example, consider a hard drive that has an MTBF of 100,000 hours (11 years), and a 5-year intended useful life. If you have 1,000 of these drives, you can expect, on average, one to fail every 100 hours within the first five years. After that, all bets are off.
So not only does a 100,000 hour MTBF not mean you'll get 11 years, you're lucky (or, more precisely, not unlucky) if you get 5 years.
As many others have said, if you intend to keep it, back it up. Every drive is only guaranteed to work until it fails.
IBM once described it this way:
http://web.archive.org/web/20001202154100/http://
FIXME: Add a sig here
Repackaged Seasonics, not Silverstones. Need more coffee.
RAID 5 isn't generally for speed (it can be a by product). It is mostly for redundancy without wasting large amounts of storage capacity, that is why it is the defacto standard in production server environments. RAID 5 includes overhead in calculating the parity, those are not there in RAID 0+1/1+0/10.
For those of you who may not be familiar with RAID 0+1/1+0/10 it is basically a stripe of mirrors (or a mirror of stripes). Because of this it is twice as fast with half the disk space. (example 4 200GB hard drives would give you a raw data capacity of 800GB and a usable capacity of 400GB, and the disk access would be twice as fast).
RAID 5 is striping with parity, basically the data is divided into as many peices as there are drives (at least three) and then parity is calculated (basically redividing the data to ensure that there is an overlap of data to allow for rebuilding of the array in the event of a single drive failure. with 4 200GB drives you would have a raw data capacity of 800GB and a usable capacity of 600GB.
$diff terrorists hippies
$
$rm -rf *terrorists *hippies
I'm a Mac user and my hard drives have lasted.... Umm Forever(+5 years)
Wait scratch that, One drive died in a power surge when I yanked the cords
to save the computer from a burning building (don't ask)
But I read about windows users replacing drives every year or so...
Now Apple uses IBM Drives and I've always upgraded with the same
are they just better or is HFS HFS+ just that much gentler on the disks than FAT16 and FAT32
There's little point in getting rid of a working hard drive unless you've got nowhere to put it. Even if you don't trust it not to fail in the near future, you can still use it to make some extra backups. Just make sure you make sufficient backups of important data.
At one naive point in my life, I believed that SMART would actually serve its intended purpose and warn me when my hard drive was going to die. SMART doesn't always work! The drive started making funny noises months before SMART even noticed.
ttuttle is a rankmaniac
I have an IBM "Deathstar" from 2001 or so, and amazingly, it's still working! Over a year ago it started making a grinding noise, but so far it hasn't crashed and burned. And the grinding noise disappeared months ago. Weird.
Storage Review has had a Hard Drive Reliability Database for quite some time now. I've listed all my drives there (Not only the ones that fail). I recommend you check it out :
g in
http://www.storagereview.com/map/lm.cgi/survey_lo
All I have to say is just make sure you buy a good quality drive. Just remember you get what you pay for. WD's get really hot and don't last as long but Seagate's stay relatively cool and I have had a seagate for about 6 years now.
You sure about that? In every production server environment I've ever worked in, downtime is a helluva lot more expensive than disks.
hang brain.
Yeah, I am sure about that. You are right downtime is more expensive than disks, however I never said it wasn't. I am having trouble understanding what point you are trying to make with your post.
RAID 5 does not equal downtime. When there is a failure in a RAID 5 array there will be no downtime if the admin is paying attention and replaces the failed drive (allowing the array to rebuild itself) before a second drive fails.
If I completely missed the point of your post please be a little more succinct in your future posts. For example if you are inferring that a RAID 5 array will have more downtime than a different type then just say that, and give some examples as to what would be a better solution and why.
$diff terrorists hippies
$
$rm -rf *terrorists *hippies
How is there no downtime? If one of the non-parity drives fails, don't you need to rebuild the array before it can be used, or do controllers allow the reading of it (or even writing) while rebuilding one of the drives?
I typically don't keep a hard drive running longer then 4 years. (After suffering a HD crash months after I rebuilt a laptop, I don't plan on keeping a laptop drive running longer then 2 years.) After 4 years or so, usually a $100-$150 drive is so much bigger that it's stupid NOT to upgrade. Don't put the old drive on the shelf and rely on it as a backup; in my experience they don't work when you plug them in a few months later.
For my important data, I keep it in Perforce (a source code control program) using an old 400mhz machine as my server. It has an automated XCOPY of the repository on a nightly basis to a second hard drive. This made recovery of my important data very quick when my laptop's hard drive died.
No, I will not work for your startup
For the record, with a good controller (or software), a 4-drive RAID 1+0 is 4x read, 2x write, not simply 2x.
Half the data is written to each stripe simultaneously, for 2x write speed, and data can also be read back twice as fast for the same reason. However, when reading half of each of those stripes can be read back from either of the stripe's MIRRORS simultaneously, resulting in 2 (striping) x 2 (mirroring) = 4x read speed.
Random and weird software I've written.
If one of the non-parity drives fails, you can calculate what was on it from the contents of the other non-parity drives + the contents of the parity drive. Any controller can do this; a good controller can do this at the same speed that it could have read the data from the disk.
And as a technical point, the parity information is spread out across all the disks for better performance. This doesn't change how RAID-5 works on a conceptual level, though.
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
My storage and speed requirements are what prompt me to replace hard drives.
I use a 200GB 8MB Cache SATA drive as my system drive. That will get replaced within six months with a 16MB cache 320GB or larger drive. Since I am a gamer, my games only(literally, it is mapped to C:\Games) drive is a 150GB Raptor.
For storage, I have a 2TB Array(8x WD 320GB RE 8MB Cache) on a separate fileserver. I use the HighPoint RocketRaid 2320 controller for it and have been very happy with it. (Great Linux support, E-mail alerts to my cell phone and gmail if any problems occur) I have been so happy with the performance of the fileserver, than I am building another separate fileserver(Well, that and I only have 60GB free on it). Using 8x Seagate 7200.10 16MB Cache drives and the HighPoint RocketRaid 2320. Critical data(about 8GB) is mirrored on my main machine, the fileserver, and DVD. All other data is considered expendable. Another data protection mechanism I use is that the only method of writing through the array is from the fileserver itself. File permissions are set to deny write access to all data on the array from any user except myself. I have an upload directory on the system drive whose folder structure mirrors the array's. I simply drop the files I want to upload in the appropriate folder, and execute my move script from my SSH session. This means that the only way for a virus or malicious user to mess with my data is to know my username and password or know the root password.
The data can be accessed the same as before, the only difference is the performance hit from the reduced number of drives, then when a new drive is added the array will automatically rebuild the drive in the background. This will have a performance impact as well, however the server is still up and accessible, although not in its top form. The only time you have an actual failure is if there is more than one drive that fails (unless you have hot spares built into the array) or if you have a raid controller fail, then there is an outage until the controller is replaced (at which point the controller would read its configuration from the drives back into the controller, a very minimal amount of time). However a RAID controller failure can happen with any sort of RAID, not just RAID 5.
All of this I am assuming you are using a server grade raid controller, I cannot speak for some of the cheaper desktop RAID controllers.
$diff terrorists hippies
$
$rm -rf *terrorists *hippies
Whatever you do, do not use RAID 0; especially not with the old drive. Performance gains are non-existent (even raw bandwidth gains are useless) and data loss risk goes up big time.
Most of my drives are older SCSI drives (some 50-pin Ultra-SCSI, others 68-pin UW), mainly Seagate 2/4/9GB drives and IBM 18GB drives, and the vast majority of them have been running almost continuously (powered up, not necessary being used) for the past several years.
I think I've had one Barracuda die during that time (a 4.5GB U-SCSI drive I'd picked up on eBay for very little which used to be used on an AOL data server).
Mainframe/UNIX Bit Twiddler and long time Windows/Linux Hobbyist.
The Theorem Theorem: If If, Then Then.
I mainly work with HP Proliant servers and RAID arrays. It seems that out of the typical 4-5 disk RAID5 array, I will typically average about a drive failure per year (per server.)
I'm talking about RAID 5 vs. RAID 1*.
RAID 5 might be fine for bargain-basement web hosts and your home file server, but no production environment that I've ever had the pleasure of working in has ever run any kind of RAID 5. Even the small risk that you might lose the parity if a disk goes down during a write is not acceptable, and rebuild time for an "ideal" failure makes it even less so. We either run straight up RAID 1 (with hot spares) for systems that aren't I/O bound, or 1+0 when disk performance is an issue.
hang brain.
My rule is that I replace BOTH of my backup hard drives when they get full, which usually takes about 2 years. And I always replace them with ones that are twice as big. I also ONLY buy drives that have at least a 3 year warranty.
Life is too short to drink bad wine!
Seasonics are great, so quiet you can't tell that they're turned on and nice stable voltage.
Well then I suppose we will have to agree that we have never worked in the same places. In all of my time I have never experienced data loss as a result of any sort of failure of a RAID 5 array.
$diff terrorists hippies
$
$rm -rf *terrorists *hippies
Any power supply worth it's salt will have active PFC. Any power supply with a voltage selector switch on the back of it is, by definition, not an active PFC power supply. Active PFC power supplies can automatically correct for AC input voltage, and thus is capable of accepting a full range of input voltages. In other words, they can protect your equipment from damaging brownouts and voltage surges.
Actually, from an engineering standpoint it's exactly as you describe. It's most likely to break right away than at any other particular time, assuming a randomly distributed set of flaws, which is common. Most electronics don't wear out so much as they fall victim to chance power and temperature fluctuations. Robustness will vary wildly in manufacturing, so any item that makes it past the beginning is pretty likely to last a while.
That's not true of things that are engineered to wear out (like tires or brake pads, but also like the bearing and motor in a HD) but in a HD the parts that wear out last quite a while, random failures are a more likely cause.
So I'd keep your old drives - and I'd keep redundant data; you shouldn't trust ANY drive.
But if you're throwing out those old 200 GB drives - or really anything over 10GB - I'd love to have them!
Looking for freelance Actionscript (Flash/Flex) or ColdFusion work and/or freelance developers. Email me, put Slashdot
To make an informed choice, I would check the HD first. First check on the age. It may be a replacement drive for the machine and not original. Many manufacturers will allow you to check on the warranty status on their website. You'll need the serial number of the drive. If it's still under warranty then that's a plus. Next run the manufacturer's diagnostics to see if there are any glaring issues like SMART errors. If it's all good then I would more than likely trust the drive. However, you should perform periodic backups on all machines. My 2 cents.
Well, there's spam egg sausage and spam, that's not got much spam in it.
Buy your hard drives OEM from your local beige-box builder instead of from a retailer like Future Shit or Worst Buy, and you get a 3 t 5 year OEM warranty instead of the stupid 1-year warranty, and you pay a bit less.
Ever had any issues with RMAs? Also, what do you ship the drives back in for the RMA? Seems like all of the OEM drives I've seen lately are only packed inside either a plastic tub or a static bag wrapped in bubble wrap.
How picky are the RMA folks about how you pack the drive for shipment back to them?
Wolde you bothe eate your cake, and have your cake?
For the record, with a good controller (or software), a 4-drive RAID 1+0 is 4x read, 2x write, not simply 2x.
Theoretically that's true... but overly optimistic. The reality is more that with a 4-drive RAID10 you'll see somewhere between 2.0x and 3.0x read speeds over a single drive. (And with 2.6.17 Linux, I have yet to see anything better then the low-end of that scale for 4-disk or 6-disk RAID10 arrays.)
Although I still plan on doing some more testing using 3-disk RAID1 arrays (3 disks, mirrored at the same time). Which is a nice trick that mdadm lets you accomplish.
Wolde you bothe eate your cake, and have your cake?
I replace a drive at the first sign that anything is odd a new noise, errors, or just slowness from a given drive. I have almost never had the "smart" system catch a drive before it has failed. I went thought a month of bad hard drive failure. I lost 6 250 gig hd's They were all purchased at about the same time and they all started failing about the same time. Drive under in order of reliably based on drive site. Drives under 120gig I've never had any failures with drive over 200 gigs. the failure rates I have seen are 1 in 6 after 1 year of service. Maxtor 250's are some of the worse drives on earth! a 100% failure rate in 18 months After converting my 6tb array over from those to seagates the failures have stopped.
failure rate depends on the conditions in which the hdd is placed in.
a friend of mine has a raid system (5 and 10) for his video storage. it used to happen that it gets an average failure of around 1 per month (in a bunch of around 20 drives). (raid is *not* is substitute for backup. it gives an added protection for a drive failure.) we live in the tropics and the ambient temp is high. i suggested to air condition 24 hours and it reduced the failure rate to around 1 every six months. note these are just desktop drives and not enterprise drives.
in another scenario, we have desktop drives running on some low end servers for years (more than 3) and we have not a single failure. of course it is placed in a datacenter.
i believe that failures of drives are caused by:
varied temperature gradients during operation (expansion and contraction)
high temperature operation (causing premature wear of components)
bad power supply and power source (this one zaps the electronics) and i suggest to get a really expensive server class psu with double conversion ups and some electrical filtering systems (if you really value your data)
dust (causes overheating)
on and off operation (if you run it 24 hours, don't turn it off)
the moral of the story: hdd fails!
the only solution is to backup backup and backup your data at different sources. personally, i duplicate my important data across two computers, a usb hdd, usb flash (for those really critical files), web drive (internet.)
Live your life each day as if it was your last.
Funny thing: I had been running LVM with no RAID for a few years with no problems when one day I said to msyelf, "You know what? This just doesn't feel safe. If one of my drives goes splat, I lose everything." So I switch to LVM on top of RAID5. Wouldn't you know it, about a month later, I lose a drive. It would have been catastrophic failure. I would have probably shrieked so loud the neighbors down the block would call 9-1-1. But instead, I just said, "darn," walked down to best buy, bought a disk, plopped it in, and rebuilt the array. It's IDE, so I had to shut down the machine to do it, but that's about the only bummer about it.
At this point, I would never build another machine without RAID. Not after it saved my behind like that.
They don't grade fathers, but if your daughter's a stripper, you fucked up. --Chris Rock
If you've got your backup drive in the same tower as your primary drive, connected to the same power supply, then any electrical or heat stress that damages your primary drive will also damage your backup drive. Also, it's extremely easy to upgrade your external drive when you want to, and you can easily use it to back up other computers (work or personal laptop, etc.), and if you're doing risky-to-computers things like moving, you can take the backup drive with you separately.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
I've returned a few, and never had a problem.
I buy a bunch of cheap sponges (12 in a pack at the local Dollar Store), and pack them in that. Thenk keep the box they ship the return drive in for next time ...