Home / server / Questions / 110415
Accepted
chris
Asked: 2010-02-08 07:34:15 +0800 CST

Desktop SATA drives in SATA <-> FC array

  • 772

Let's assume you've got a box like one of these with space for 24 SATA disks.

What are the best bits of advice for deploying this?

For instance, should you be greedy and go for the 1.5 or 2tb disks or are they just not reliable enough to be used in an array like this and you should stick with 640gb or 750gb disks instead?

Also, I know that FC (or generically, "enterprise class") disks have a different error recovery strategy than desktop disks. An enterprise disk will fail a read quickly and report to the controller that it wasn't able to read that block, and the RAID controller will quickly regenerate the info from the parity disk and mark the block as bad. A desktop disk, on the other hand, will try and try and try again to get the data, and in pathological cases this may cause a raid controller to fail the whole disk because the read operation times out.

So there are a couple aspects to this question:

  1. What's the best sort of disk to get today? (ie specific disks on the market in Feb 2010)

  2. Generically, what should someone look for when trying to buy something like this that kinda walks the line between enterprise and consumer?

  3. Lastly -- is there anything that can be done with current "consumer" disks to make them more suitable for array use? IE can you use a SMART configuration to change the error recovery strategy used by the disk?

Thanks!

hardware-raid sata

2 Answers

  1. Best Answer

    Okay -- this is what I've found so far:

    The main concrete difference between an "enterprise" and "consumer" disk is the strategy used by the disk to recover from read/write errors.

    • consumer=try really hard, even if it takes a minute, to get the data back

    • enterprise=give up fast and assume the controller will get the data from the parity disk or mirror

      Samsung calls this "Command Completion Time Limit" (CCTL) RAID Error recovery and Western Digital calls this "Time Limited Error Recovery" (TLER) and Seagate calls this "Error Recovery Control".

    As far as I can tell:

    • samsung: enterprise drives have CCTL, consumer drives don't.
    • WD: There is a utility people talk about on forums that enables TLER on their consumer drives. The forums typically say things like "you have to get it from WD directly, who denies that it exists, or you can go to this rapidshare link that may or may not still work." Hmm. Other forums report that it is now impossible to change the TLER setting on any newer WD disks and that using the TLER.exe utility may break disks or cause male pattern baldness.
    • Seagate: It looks like some drives have the "ERC" register, some don't; if the register is there it looks like you can set it with a SMART management utility like hdparm under linux. I haven't tried setting the ERC register -- none of the disks I have at my disposal have the register.

    Lastly, I suspect but am not sure that all of these are implementations of the SMART feature "Error Recovery Control" (see page 34 of this PDF). If anyone knows for sure, please comment.

    I have no specific knowledge of which kinds of disks are more or less reliable. I suspect that unless I were buying tens of thousands of disks, I wouldn't ever be able to tell which are more reliable (consumer or enterprise). As always, I'd go with superstition, warranty length, and number/quality of newegg rants.

    I mostly got this information from the linked articles above and these forums/articles (that hopefully will remain where they are until the end of time):

    http://forums.seagate.com/t5/Internal-ATA-and-Serial-ATA/RAID-Issues-ERC/m-p/25751

    http://www.anandtech.com/storage/showdoc.aspx?i=2922

    http://www.gohighvoltage.com/forum/index.php?topic=50.0

    http://hardforum.com/showthread.php?t=1285254

    • 2

  2. Firstly that disk array will take regular disks but take into account that most consumer drives are not meant to deal with an enterprise-type work load such as working >40% of the time and being switched on 24/365.

    That said I'd have no real problem running one with the latest 1.5GB drives but probably wouldn't go the whole hog with 2TB drives simply because the 1.5's are more stable as they're older and will handle the potential heat build-up better.

    In terms of 'best' disk, well I MUCH prefer SAS disks, they're just designed and built better - forget any performance gain I just see them failing less and handling constant use better than even enterprise-class SATA disks.

    Whatever you choose just make sure you have at least 2 spare disks sat right next to the array and that you have disk-failure/pre-failure alerting switched on and tested - most array-based data loss is due to inaction due to lack of alerting.

    Oh and I've no idea about your last question other than go for disks with less platters as they handle heat-variance better.

    • 1