Slashdot Mirror

A double disk failure may be very unlikely, but a disk failure combined with a read-error during rebuild isn't...

Nice little blog about that (from a manufacturer) ;
http://blogs.netapp.com/dave/TechTalk/2006/03/21/E xpect-Double-Disk-Failures-With-ATA-Drives.html

I just got back from the results meeting for an Oracle RAC evaluation, done on Solaris 10, with various filesystem options. The O/S support was no big deal, but the DBAs had a really rough time of it, even with Oracle, Sun, and Veritas consultants along for the entire ride. One DBA checked out halfway through the three-month process, and the replacement couldn't go a single day without mentioning how simple RAC setup and configuration is with Red Hat. The project is on hold, pending better explanations from Oracle as to why it took so long just to test it.

Even though Oracle and Sun seem to go together--at least, in the minds of IT management--there are several documented gaps, e.g., Oracle on Solaris with NAS storage is painfully slow without tuning. Fortunately, there's a whitepaper that covers this particular case pretty well. Not so for every deficiency in the Oracle/Solaris combination.

Surely Oracle/Linux has quirks of its own, but like the ubiquitous Oracle/Solaris environment, these will surface with time + a growing installed base. The firm I'm supporting is very risk-averse, e.g., many meetings must precede even the evaulation of new technology, with many more meetings (and man-hours) before the pilot. I don't see them even testing Oracle on Linux for another year.

By the same token, how many early adopters for a DBMS (even one with ludicrously huge market share) with its own O/S?

See also the Pick Operating System.

I do however see a need for super computers who need to work with filesystems spanning perhaps hundreds of disks.

Super computers? Once, maybe - not today, and not for the last decade or so. There are a bunch of companies (I'm working for one of them, now) that will quite cheerfully sell you a storage system that spans hundreds of disks. Assuming your OS won't flake out when it sees a 500+ TB volume, you could mount it on your desktop, if you want. There's absolutely no need to conflate processing power (super computers) with storage capability (NAS, SAN, etc.)

I do however see a need for super computers who need to work with filesystems spanning perhaps hundreds of disks.

Super computers? Once, maybe - not today, and not for the last decade or so. There are a bunch of companies (I'm working for one of them, now) that will quite cheerfully sell you a storage system that spans hundreds of disks. Assuming your OS won't flake out when it sees a 500+ TB volume, you could mount it on your desktop, if you want. There's absolutely no need to conflate processing power (super computers) with storage capability (NAS, SAN, etc.)

My problem this time was that I BELIEVED what the customers were saying. Just because every major EDA and FPGA company issued press releases supporting LPM doesn't mean that they actually want anything to do with it (it's basically now an Altera specific format). Learning the difference between what a customer will buy, and what he says he will buy is key.

http://www.netapp.com/support/ues.html

All Network Appliance software license terms and conditions specify a "license to use," therefore software cannot legally transfer from one owner to another. Anyone purchasing used hardware equipment must also purchase new software licenses directly either from Network Appliance or from an authorized Network Appliance reseller. Software includes all protocols as well as streaming licenses, Snap products, and other software. Anyone trying to sell you "used software" would be violating the terms of the license. Support contracts such as warranty and maintenance agreements are also non-transferable.

However, even a low-end system would be pretty pricey for home use, and you'll probably balk because it's many times more expensive than a cheap and easy linux box with a samba, some cheap disk drives, and a raid controller. However, the features and performance that a Netapp server provides more than make it for it.

Still, you might be better off trying to find an used, older Netapp model that's still supported and buying it second-hand. You'll still have to license the software, but you might be able to find a system that's more in-line with your needs.

Bruce

Once you understand that RAID is reliability strategy and are prepared to have appropriate backup measures in place, then RAID 5 becomes an attractive option for the home network. I've recently looked at several options.

• LaCie Biggest Disk - Cheap but of questionable reliability. Since RAID systems should be reliable above all else, I would rule this out.
• Buffalo TeraStation: An interesting product but again reviews are pretty mixed.
• FirewireDirect Vanguard V5: Solid offering from a company that focuses primarily on larger scale storage solutions.
• NetApp: A well regarded product primarily aimed at corporate users.

In my case, a three disk RAID 1 solution proved more appropriate than RAID 5. I value high reliability on the home system and wanted to use a rotating third disk as a backup in the event of catastrophic data loss (e.g. house burns to ground). FWIW, I also use a DAT for differentiatial backups. For many users this may be overkill -- sacrificing three disks plus fixed hardware costs to greatly reduce potential data losses -- but for priceless coding projects and digital pictures, this might be good for you as well.

For some users working with video or having large audio collections, much larger disk systems may be desired. First make sure that you have an appropriate mechanism for backing up a terabyte or three. Then, the Vanguard V5 may be an excellent solution if the $2-3k price is acceptible.

Open Firmware does not work on Intel processors.

The NetApp FAS900 filers, and most earlier NetApp x86 machines, use Open Firmware (the exceptions were the machines, mostly NetCache machines, using standard Intel boxes OEM'ed). Now, that was a port of the Firmworks OpenFirmware code to x86, rather than a version of Apple's independently-implemented Open Firmware implementation, but there's nothing technical that prevents Open Firmware from running on x86.

> Nobody uses nfs on new implementations today anyway

Oh really? NetApp Customer Stories

There's several (mostly unknown) companies and orgs not even on this list with ongoing IP SAN projects using NFS as their storage protocol.

Compared with the alternatives, such as IBM's entry-level system http://www-132.ibm.com/webapp/wcs/stores/servlet/C onfiguratorDisplay?storeId=1&catalogId=-840&langId =-1&site_type=public&oiId=null&currency=USD&base=1 722-6LU&x=11&y=12/, Dell http://www1.us.dell.com/content/products/compare.a spx/das_storage?c=us&cs=04&l=en&s=bsd/ or NetApp http://www.netapp.com/, add the cost of the redundant power and cooling, and ignoring the cost of trying to maintain a Linux system alongside their OS-X, Apple's price is not bad for the storage supplied.

Sometimes, it really is time to call in professionals, and spend the money to do it right. I've built some storage solutions on my old job, to support our compute clusters. They worked from an OS point of view, but even buying what were supposed to be well-regarded components resulted in more downtime than we considered acceptable, mostly from heat. In the end, as soon as we had the money, it was entirely worth it to be able to call IBM, buy their servers for disk storage, and get the better engineering in terms of cooling, drive access, and remote management, as well as the three year warranty with onsite service.

It is highly unlikely that you can screwdriver together a system with as good of airflow as major vendors who employ real engineers. As the poster said, it's a business, therefore when your data matters is not the time to hack something together with parts from NewEgg.

I know they're late to the party, but the FAS3000 line supports the new shelves with SATA drives. This drops their cost down dramatically (over 50% reduction), with the same unusually high NetApp standards. We're upgrading our F820c cluster with FAS3050c heads, and adding some SATA storage for new projects. I'd have to agree with the parent that NetApp is definitely a good solution, though none of their individual filer heads, or even two head clusters, can scale this far up. Their single volume limitation is still at 17.6TB.

In the past, our management splurged and bought a Network Appliance. Two, actually; we have an F880 and an F740. The 740 is pretty much defunct right now; it's only got about 300GB of disk and we use it to house static application installs. The 880 is more robust and therefore more efficiently used, but only has 2.2TB of (insanely reliable) disk. Disk that costs about $20,000 per additional TB.

Last year, we learned that Xyratex (the company that makes the disk shelves for NetApp) has started selling SATA-based disk arrays. Right now, I believe they only support 400GB SATA drives in a 16-drive chassis, but support for 500GB drives is supposedly right around the corner. A fully-populated chassis with 400GB drives will yield about 4.8TB of usable space. We have purchased five head units (about $20,000 full of disk) and one shelf (about $15,000 full of disk). Each unit is expandable to 7 shelves (including the head), which yields over 32TB of usable disk. I don't know what your budget is, but $110,000 is pretty reasonable for over 32TB. Admittedly, you could buy 32 stripped-down Dell Dimension 4700's (to get SATA) with two 500GB hard drives; install a slim OS and you could get approximately the same amount of usable space. But the reliability of the Xyratex has been far greater than the reliability of the Dell machines we've purchased in recent years.

It's ironic that you brought up the NeoPath File Director. We're going through the trials and tribulations of installing a clustered pair of them right now. We've had some difficulty in getting them set up, but it seems like they'll do the trick when we get them going. The MSRP on the cluster is kind of high, but talk to their sales guys if you're interested - we got almost 1/3 off the listed price. We plan on using the File Director to migrate old files from the NetApps to our Xyratex, thereby expanding our storage at $3,300 per TB instead of $20,000 per TB. I can see it working well, though, for aggregating a large number of file servers into a single virtual server.

I don't know how the File Director will interface with your operating systems. We use Veritas's Storage Foundation (~$500 per license - you should only need one) on our systems because we're primarily limited to Windows and Solaris, which have difficulty with large filesystems. Storage Foundation breaks the size limitations as well as enabling easier management of your volumes.

I hope this helps. Good luck.

Go Network Appliance and you get the same treatment. Their hardware is rock solid and completely hands off for many months at a time. I only ever touch them to add filesystems. You get what you pay for, so remember that if you cheap out YOU have to live with it, not your boss.

there are some vendors also like netapp (http://www.netapp.com/ well known for this kind of stuff.

You should test certainly iscsi implementation on linux.

Open Firmware on x86????

Yeah - try one of these boxes, for example.

Full disclosure: I work for NetApp. In the marketing department, even... yuck!

But... one of the easiest explanations of how snapshots work is in a ten-year-old whitepaper, located here:
http://www.netapp.com/tech_library/3002.html#I34

Also... I'm not familiar with the $PWD/.gone feature described above; I'm pretty sure that must be a local hack, maybe a symlink pointing to one of the snapshots in the .snapshot directory. Out of the box, to recover a file you have deleted, you do something like this:

$ cd .snapshot
$ cd hourly.0
(or hourly.10, or weekly.4, etc. depending on which version of the file you want to recover)
$ cp myfile ~/projdir/myfile

No separate .gone directory is needed because the .snapshot directory holds all changes since the snapshot was made; including modifications to files as well as deleted files. For simplicity, it even holds the things that have NOT changed, so looking at a snapshot directory is exactly like looking at a read-only version of the real directory, frozen at a certain moment in time.

There's some marketing stuff here, but I can't find any technical guides:

http://www.netapp.com/products/software/snapshot.h tml

Basically, the Netapp takes a snapshot of your directories on a regular schedule (hourly, nightly, weekly, whereever). If you want to recover an old version of the file, it's as simple as a 'cp .snapshot/hourly.0/foobar.sh .'

Also, when you rm a file, the Netapp can put the removed file in a $PWD/.gone directory in case you removed it by accident, and removes the file later (1 day later, 7 days later, whatever you need).

Great features.

This is why you use a NetApp. Backups are atomic, guaranteeing filesystem integrity, so recovering from a restore is exactly identical to recovering from a system crash, except you have more guarantees about the state of your data. I've used NetApp snapshots (the fundamental building block upon which backups are based on a NetApp) to back up Oracle databases that were under heavy read/write load, and restored backups of same. No worries.

...and as we all know, a server host can write about 150MB/sec...

Pretty much all our primary data storage is on Network Appliance filers. All these systems are configured with hourly and nightly snapshots on the filesystem.

We also snapmirror all the changes from primary storage every hour or so to 24TB of (relatively) cheap secondary storage (One of these )

Some of the primary storage is at remote sites with slower WAN links so we just throttle the snapmirror down.

The secondary storage is hooked up to a tape library so we can send long term backups offsite.

Pretty much all our primary data storage is on Network Appliance filers. All these systems are configured with hourly and nightly snapshots on the filesystem.

We also snapmirror all the changes from primary storage every hour or so to 24TB of (relatively) cheap secondary storage (One of these )

Some of the primary storage is at remote sites with slower WAN links so we just throttle the snapmirror down.

The secondary storage is hooked up to a tape library so we can send long term backups offsite.

I think the sheer amount of projects and their sizes presents a lot of the problem. At my university I've been given a lot of old hardware to create our internal mirror. 500GB goes fast when you get all of Debian or Fedora. I have to constantly evaluate what is important, since I don't have the funds to throw more storage at the box.

If I could just get my hands on our new NetApp...

If they could come up with a 350MB or 3.5TB removable drive at a reasonable price, they might have a viable idea. But as it stands, with their reputation in tatters, I doubt this will go very far.

Sure they can. I've seen plenty of arrays that can keep up with 250MB/s. In fact anything attached to a decent database server or used as central storage for a number of servers better be able to keep up that kind of throughput. For instance Netapp's FAS960C cluster solution can push 2.5GB/s on a synthetic benchmark

Actually this thing would have melted down and cried in the face of Weta's storage needs, just because it's dense doesn't mean it's fast. You need lots of spindles, and 4 doesn't count. The average FibreChannel disk is good for maybe 200-300 ops per second, so you really want lots of 'em.

They use a mix of NAS and DAS storage, I belive the DAS consists mainly of SGI stuff using XFS and CXFS, while the NAS stuff is all NetApp Filers which can have hundreds of disks

NetApp PR

SGI PR

B

Network Appliance is a company that provides NAS products. I was answering the question as to whether or not there was a NAS solution that reliably did NFS and CIFS. to be precise :)

What you're asking for is a SAN.

I just installed a Network Appliance FAS250 in my server room. It speaks CIFS, NFS, and iSCSI.

By the way, you're wrong... Oracle will run perfectly using CIFS shares (I'm running it now, and have been for the past few months), and NetApp has plenty of documents in their tech library showing all the different ways to use attached storage with Oracle and many other pieces of software.

With respect to speed, it really depends on the network infrastructure. I've got a Cisco GigE switch attaching 6 machines directly to a GigE port on the NetApp Filer. It is literally twice as fast than the directly attached RAID 5 (caching, etc.) arrays that it replaced.

I think that Microsoft Exchange can be installed to a CIFS share, but if not, you should look at iSCSI. My company uses Lotus Notes 4.6.7 (sweet, merciful Christ, please put me out of my misery), and it works great from a CIFS share on the NetApp.

Microsoft has a free iSCSI Initiator for Windows that will mount an iSCSI device just like any other SCSI drive in Windows. You can find several iSCSI targets for linux here.

I have about 50 Mac's on our network (graphics department) that needed to talk with the new filer. Instead of installing a klugy piece of software to make the OS9 Macs talk to the SAN at $150/seat, I installed a linux box using samba to talk to the SAN through CIFS and netatalk (AppleTalk for linux) to re-share out the samba mounts. Becides some quirks (Mac's don't see the linux gateway in the AFP browse list, but can connect directly through IP), it works rather well.

Look at iSCSI, it does exactly what you're looking for.

What you're asking for is a SAN.

I just installed a Network Appliance FAS250 in my server room. It speaks CIFS, NFS, and iSCSI.

By the way, you're wrong... Oracle will run perfectly using CIFS shares (I'm running it now, and have been for the past few months), and NetApp has plenty of documents in their tech library showing all the different ways to use attached storage with Oracle and many other pieces of software.

With respect to speed, it really depends on the network infrastructure. I've got a Cisco GigE switch attaching 6 machines directly to a GigE port on the NetApp Filer. It is literally twice as fast than the directly attached RAID 5 (caching, etc.) arrays that it replaced.

I think that Microsoft Exchange can be installed to a CIFS share, but if not, you should look at iSCSI. My company uses Lotus Notes 4.6.7 (sweet, merciful Christ, please put me out of my misery), and it works great from a CIFS share on the NetApp.

Microsoft has a free iSCSI Initiator for Windows that will mount an iSCSI device just like any other SCSI drive in Windows. You can find several iSCSI targets for linux here.

I have about 50 Mac's on our network (graphics department) that needed to talk with the new filer. Instead of installing a klugy piece of software to make the OS9 Macs talk to the SAN at $150/seat, I installed a linux box using samba to talk to the SAN through CIFS and netatalk (AppleTalk for linux) to re-share out the samba mounts. Becides some quirks (Mac's don't see the linux gateway in the AFP browse list, but can connect directly through IP), it works rather well.

Look at iSCSI, it does exactly what you're looking for.

What you're asking for is a SAN.

I just installed a Network Appliance FAS250 in my server room. It speaks CIFS, NFS, and iSCSI.

By the way, you're wrong... Oracle will run perfectly using CIFS shares (I'm running it now, and have been for the past few months), and NetApp has plenty of documents in their tech library showing all the different ways to use attached storage with Oracle and many other pieces of software.

With respect to speed, it really depends on the network infrastructure. I've got a Cisco GigE switch attaching 6 machines directly to a GigE port on the NetApp Filer. It is literally twice as fast than the directly attached RAID 5 (caching, etc.) arrays that it replaced.

I think that Microsoft Exchange can be installed to a CIFS share, but if not, you should look at iSCSI. My company uses Lotus Notes 4.6.7 (sweet, merciful Christ, please put me out of my misery), and it works great from a CIFS share on the NetApp.

Microsoft has a free iSCSI Initiator for Windows that will mount an iSCSI device just like any other SCSI drive in Windows. You can find several iSCSI targets for linux here.

I have about 50 Mac's on our network (graphics department) that needed to talk with the new filer. Instead of installing a klugy piece of software to make the OS9 Macs talk to the SAN at $150/seat, I installed a linux box using samba to talk to the SAN through CIFS and netatalk (AppleTalk for linux) to re-share out the samba mounts. Becides some quirks (Mac's don't see the linux gateway in the AFP browse list, but can connect directly through IP), it works rather well.

Look at iSCSI, it does exactly what you're looking for.

What you're asking for is a SAN.

I just installed a Network Appliance FAS250 in my server room. It speaks CIFS, NFS, and iSCSI.

By the way, you're wrong... Oracle will run perfectly using CIFS shares (I'm running it now, and have been for the past few months), and NetApp has plenty of documents in their tech library showing all the different ways to use attached storage with Oracle and many other pieces of software.

With respect to speed, it really depends on the network infrastructure. I've got a Cisco GigE switch attaching 6 machines directly to a GigE port on the NetApp Filer. It is literally twice as fast than the directly attached RAID 5 (caching, etc.) arrays that it replaced.

I think that Microsoft Exchange can be installed to a CIFS share, but if not, you should look at iSCSI. My company uses Lotus Notes 4.6.7 (sweet, merciful Christ, please put me out of my misery), and it works great from a CIFS share on the NetApp.

Microsoft has a free iSCSI Initiator for Windows that will mount an iSCSI device just like any other SCSI drive in Windows. You can find several iSCSI targets for linux here.

I have about 50 Mac's on our network (graphics department) that needed to talk with the new filer. Instead of installing a klugy piece of software to make the OS9 Macs talk to the SAN at $150/seat, I installed a linux box using samba to talk to the SAN through CIFS and netatalk (AppleTalk for linux) to re-share out the samba mounts. Becides some quirks (Mac's don't see the linux gateway in the AFP browse list, but can connect directly through IP), it works rather well.

Look at iSCSI, it does exactly what you're looking for.

What you're asking for is a SAN.

I just installed a Network Appliance FAS250 in my server room. It speaks CIFS, NFS, and iSCSI.

By the way, you're wrong... Oracle will run perfectly using CIFS shares (I'm running it now, and have been for the past few months), and NetApp has plenty of documents in their tech library showing all the different ways to use attached storage with Oracle and many other pieces of software.

With respect to speed, it really depends on the network infrastructure. I've got a Cisco GigE switch attaching 6 machines directly to a GigE port on the NetApp Filer. It is literally twice as fast than the directly attached RAID 5 (caching, etc.) arrays that it replaced.

I think that Microsoft Exchange can be installed to a CIFS share, but if not, you should look at iSCSI. My company uses Lotus Notes 4.6.7 (sweet, merciful Christ, please put me out of my misery), and it works great from a CIFS share on the NetApp.

Microsoft has a free iSCSI Initiator for Windows that will mount an iSCSI device just like any other SCSI drive in Windows. You can find several iSCSI targets for linux here.

I have about 50 Mac's on our network (graphics department) that needed to talk with the new filer. Instead of installing a klugy piece of software to make the OS9 Macs talk to the SAN at $150/seat, I installed a linux box using samba to talk to the SAN through CIFS and netatalk (AppleTalk for linux) to re-share out the samba mounts. Becides some quirks (Mac's don't see the linux gateway in the AFP browse list, but can connect directly through IP), it works rather well.

Look at iSCSI, it does exactly what you're looking for.

What you're asking for is a SAN.

I just installed a Network Appliance FAS250 in my server room. It speaks CIFS, NFS, and iSCSI.

By the way, you're wrong... Oracle will run perfectly using CIFS shares (I'm running it now, and have been for the past few months), and NetApp has plenty of documents in their tech library showing all the different ways to use attached storage with Oracle and many other pieces of software.

With respect to speed, it really depends on the network infrastructure. I've got a Cisco GigE switch attaching 6 machines directly to a GigE port on the NetApp Filer. It is literally twice as fast than the directly attached RAID 5 (caching, etc.) arrays that it replaced.

I think that Microsoft Exchange can be installed to a CIFS share, but if not, you should look at iSCSI. My company uses Lotus Notes 4.6.7 (sweet, merciful Christ, please put me out of my misery), and it works great from a CIFS share on the NetApp.

Microsoft has a free iSCSI Initiator for Windows that will mount an iSCSI device just like any other SCSI drive in Windows. You can find several iSCSI targets for linux here.

I have about 50 Mac's on our network (graphics department) that needed to talk with the new filer. Instead of installing a klugy piece of software to make the OS9 Macs talk to the SAN at $150/seat, I installed a linux box using samba to talk to the SAN through CIFS and netatalk (AppleTalk for linux) to re-share out the samba mounts. Becides some quirks (Mac's don't see the linux gateway in the AFP browse list, but can connect directly through IP), it works rather well.

Look at iSCSI, it does exactly what you're looking for.

"Whatever happened to having things small and quiet?"

Small and quiet are still with us.

These numbers are great, but only tell us a little about reliability or "real world" performance. When I did testing on these file system I used all the benchmarks here, plus a benchmark called postmark. This benchmark utility was released into the public domain by Net App and has to be one of the better "real world" benchmark suites.

The problem that we had with JFS during testing is that we had kernel panic with very large files. Thus we chose XFS - which has done an excellent job. I'm sure glad that the XFS file system has been merged into the 2.6 kernel, no more patching the 2.4's!

For more benchmarks on other file systems using postmark check out This

These numbers are great, but only tell us a little about reliability or "real world" performance. When I did testing on these file system I used all the benchmarks here, plus a benchmark called postmark. This benchmark utility was released into the public domain by Net App and has to be one of the better "real world" benchmark suites.

The problem that we had with JFS during testing is that we had kernel panic with very large files. Thus we chose XFS - which has done an excellent job. I'm sure glad that the XFS file system has been merged into the 2.6 kernel, no more patching the 2.4's!

For more benchmarks on other file systems using postmark check out This

I'll let the good folks at netapp do a description, they are far better at it than I. Descrition

• Weta Digital develops 'Lord of the Rings' visual effects on IBM Linux Intellistations
  
• WETA's Oscar-winning LOTR Digital animation
  
• Network Appliance Chosen for The Two Towers
  

Absolutely, and the world would be a more interesting place if more ISPs had implemented in-house CDNs rather than leaving it at basic HTTP proxying.

Like you say, it is relatively easy to implement some sort of DNS proximity testing, possibly using BIND views (or a box from Alteon/Cisco/whoever) and some standard caches (Squid, NetCache etc.). Streaming media can also delivered in this manner to save on core network bandwidth using application-layer stream splitting from people such as Network Appliance.

OK, I work in a similar environment - a little looser than yours ( we do trade paper processing for banks).

However, what was done in the article, is just a homegrown implementation of what NetApp is doing. You use the IDE Raid for nearline storage. Remember, doing backups takes time and bandwidth, databases slow their response when exporting, etc.

So, instead, take your snapshot of data at disk copy speeds, then use the IDE as a staging to tape. And it is a nearline storage for fast restores of you last snapshot.

All of this, is just a revisiting of the old concept of HSM.

OK, I work in a similar environment - a little looser than yours ( we do trade paper processing for banks).

However, what was done in the article, is just a homegrown implementation of what NetApp is doing. You use the IDE Raid for nearline storage. Remember, doing backups takes time and bandwidth, databases slow their response when exporting, etc.

So, instead, take your snapshot of data at disk copy speeds, then use the IDE as a staging to tape. And it is a nearline storage for fast restores of you last snapshot.

All of this, is just a revisiting of the old concept of HSM.

but on IA32 (well, on any architecture, but the current problem only really applies to cheap-and-popular IA32),

...or an apple XServe (unless they go to, say, a PowerPC 970 in future models).

can only use up to the addressable limit.

...at any given time. An application could use a memory-mapping API such as mmap() to map pieces of a >4GB object into the address space as needed. Yes, that's rather ugly, but people have done it before, e.g. on PDP-11s back in the old days.

Another thing you can use lots of memory for is a disk cache; I know of one series of x86 machines that supports more than 4GB of memory, most of which is used for disk caching (given that all those machines do is file service).

It is not just the costs of the raw disk space, but the enormous time and expense required to maintain/backup/restore a large database. Your 25 cents will not go very far.

Also, the access times would increase with more data to churn through, causing complaints about lag. These raise their CS costs and also cause bad word of mouth on the boards.

They are trying to attract mainstream folks who have never tried evercrack and want something more than Sims Online. This means not catering to the muling that the average person would find unfair.

Our test of the promise raid under redhat linux with the "open source" drivers (2.4.19 vanilla) compared with the 3ware product were VERY different.

I don't have the exact numbers off hand, but the 3 ware product was roughly 3 times faster at reading (raid 0+1 and raid 1). The 3ware was also faster at writing albeit the numbers were much closer. The number that DOES stick in my head was the postmark benchmark from netapp we ran. The promise did 2500 files, from 2 to 200k with 500 operations in about 35 seconds. The 3ware product did the same in 12.

The moral of the story is TEST TEST TEST, these types of articles only give you an idea. Promise worked great for me personally in several applications. After testing it for a production machine at work, we went with the 3ware because the promise did not perform well for our application. Test for youself, or forever be dissapointed.

Cluge

As other posters have noted, this isn't uncommon; Oracle comes immediately to mind, for example.
But more surprising is Network Appliance's used equipment policy. You can resell your netapp equipment, but you can't transfer the license. Instead the buyer mustalso buy a new license...which coincidentally costs the same as a whole new machine!

What I want is a filesystem that supports 1)Long File Names 2)Large Files (over 4GB) 3)Journaling and 4) can be used between Linux, Mac OS X and whatever else. Unfortuantely there are currently no filesystems that mett all of these criteria.

What you want is one of these.

A few years ago I saw a neat (expensive!) disc array that could 'freeze' the disc image at a single point in time so that a backup could be taken from the frozen image. The backup software saw only the frozen image, while the rest of the OS saw the disc as normal including updates made after the freeze occurred. The disc array maintained the frozen image until the backup was complete, guaranteeing a true snapshot as at a specific instant in time.

Sounds like the Network Appliance Filer's "snapshot" feature, but less advanced. (You can also get exactly the feature described under Linux purely in software, via LVM, now.) Under the NetApp version, you gain an extra directory ".snapshot", which contains previous versions of each file. So, if you screw up editing some file (delete/corrupt it, whatever) you can just grab a previous snapshot copy. Like having a series of online backups - but without all the extra space+hardware needs. Like CVS, but without the hassle (or fine-grained control) of doing "commits". Just tell the Filer "take a snapshot now" and 30 seconds later, it's done. Or "take snapshots every hour".

Neat feature - you could almost get this using LVM under Linux, but not quite...

Having said that, if you used the Mocha with some sort of NAS device like a NetApp then you might have something. IMO, the RLX stuff is still a better solution, though. At least with those you can mirror the boot drive.

Of course, I'm just pulling all this outta the air, so take as many grains of salt as needed.

-B

Yes indeed, this is a complicated, sure-to-cause-more-problems-than-it-solves solution to a non-problem. Export filesystems read-only to your static Web servers and read-write to your back-end thinkers (DB servers, content management systems, etc).

If you're really smart, you're doing all of this on a netapp filer so that the access speed is as good as or better than local-attached storage (and, yes that's true even though it sounds wacked... it's because of thier NVRAM-based journaling filesystem for which their NFS server code is hand-tuned).

...(squid is a better solution than ANY commercial filter/proxy on the planet. oh and it happens to be FREE!)

As it happens, I am using Squid... with the SmartFilter plugin. This took a bit longer to install than Squid by itself, and is considerably more expensive ($4K/year for a 100 user license).

the porn filters that are freely available work just fine.. if they want to add other "naughty" sites, it takes exactly 30 seconds to add it to a flat-text file... even a MSCE coud do it.

Now then, where do I download equivalent lists to the couple of dozen additional categories that SmartFilter offers? Games? Drugs? Sports? Mp3? Chat? Investing? Dating? Webmail? Anonymizer? Politics/Religion? Cults/Occult? etc, etc.

paying for any type of filtering system is pure stupidity and would only be reccomended by incompetent sysadmins/netadmins.

Paying a subscription for updated filter information is not stupid if (and only if) you can trust the commercial entity that is maintaining the list to do a good job and not let any of their bias (religious, political, etc) show in how they rate sites.

I wish I could make it my job to surf for new Mp3 and porn serving web sites to add the URLs to a corporate filter list, but the reality is, paying $4K+ per year to subscribe to a list of sites/categories is cheaper and more effective than paying me to do the same job by hand (pun intended).