Barracuda XT, Barracuda, Barracuda Green, DiamondMax and other desktop SATA and IDE drives

[matt_callaghan]

As soon as I have both drives back into a RAID1 array, (http://forums.seagate.com/stx/board/message?board.id=ata_drives&thread.id=15562&view=by_date_ascending&page=1), I will be able to perform some diagnosis to help us figure out this problem.  Please note that while I've very little experience with diagnosing hard drives, I am experienced with both Windows and Linux, so hopefully this helps.  That said, I should simply need some good instructions for diagnosis and I will most certainly be able to execute the test(s).

My ultimate goal is to be able to provide Seagate with PROOF that something is defective with the 7200.11 drives which render them inoperable due to a RAID configuration.  This is assuming of course that this is the case, which based on reports from this thread, seems extremely probable.

Can someone please recommend some sort of "controlled" tests that migh thelp us narrow down the issue?

Message Edited by matt_callaghan on 10-13-2009 03:42 PM

[fzabkar]

It seems to me that the major complaint with these drives in a RAID configuration is that they take to long to come back from an error condition, in which case the RAID controller times out and drops them from the array. I suggest that you enable Error Recovery Control (ERC) in your drives and set their R/W time limits to be less than the timeout limit of your controller.

The author of HDAT2 has recently added an ERC Menu to his software and has asked people to test it:

http://www.hdat2.com/

Another thing you could check is whether Spread Spectrum Clocking (SSC) has been enabled. HDAT2, MHDD, Hitachi's Feature Tool, and Seagate's sscset.exe should be able to enable/disable SSC.

http://www.seagate.com/support/kb/disc/faq/ssc.html

http://hddguru.com/content/en/software/2005.10.02-MHDD/

http://www.hitachigst.com/hdd/support/download.htm

http://www.hitachigst.com/hdd/support/ftool_user_guide_198.pdf

Some drives support the ATA Write-Read-Verify feature set (see section 7.48.10 of the ATA-8 spec below). This feature enables the drive to automatically verify the data that have just been written. However, the improved data integrity would probably come at a measurable performance cost. On the upside, one could determine whether problems occurred during the initial writing, or later during readback.

If you suspect that bad sectors are been created during "head slap" incidents, such as when the heads dive during deceleration, then you could turn on Automatic Acoustic Management (AAM). This will reduce the aggressiveness of the voice coil's acceleration and deceleration profiles.

To test for temperature as a contributing factor, use a household fan to keep the drives cool.

If you suspect that firmware is an issue, capture the Identify Device data blocks of each drive and compare their features. You could use smartmontools for this purpose. The following command line produces a debug output:

smartctl -a r ioctl,2 your_device

smartmontools is a Linux application that also has a Windows port.

http://sourceforge.net/projects/smartmontools/files/

The Identify Device data are explained in the ATA-8 command spec (section 7.16):

http://www.t13.org/Documents/UploadedDocuments/docs2008/D1699r6a-ATA8-ACS.pdf

[Undermoose]

Ok so, end of August I buy 6 ST31500341AS drives (All 6 are 7200.11 - 1.5TB - CC1H) from Best Buy (on Sale $119 bucks that weekend) for my Asus Rampage Extreme mobo, which has the Intel ich9r drive controller with 6 ports.
 
I build a 6 disk raid10 volume (1411.11gb) and then another 6 disk raid10 volume (2780.5gb). The first takes 8 hours to build, and then the second takes another 10+ hours to build. Slick I now have two giant raid 10 volumes.
 
Everything is perfect for 6 weeks, and then this past weekend I had the joy of letting the array rebuild from a dropped disk (failed in the controller). I simply marked the disk as good and let the array rebuild (yeah this takes a while to rebuild).
 
The array finished rebuilding and everything looks fine, so then I then arrive here to check on what's what with 7200.11 and EEK! I sure hope I don't have constantly dropping out disks... OUCH!
 
 
Windows 7 RTM 64bit Ultimate with Intel 8.9.0 build 1023 Matrix Storage Manager.

[nik_0_0]

After several google searches for the many problems with my computer, first with the terms 'security options failed to create logon process' and countless others, I finally found this his with 'iastor1 time out' message. This thread describes exactly what has happened to me over the past year.

I have two Seagate Barracuda 7200.11's, and ever since about 8 months ago, VERY infrequently, my computer would crash (usually overnight, possibly due to a large amount of access) and then upon reboot the drive would be any combination of degraded, not found, not readable or simply not even there. But as some have stated, this drive reappears without a hitch after a power cycle. This behaviour was an annoyance, but was not frequent enough to warrant any complaint. However, just recently, these problems became more and more frequent, and more severe.

My computer now hangs after about an hour of use, while my mouse is still operable, everything else is compeltely non responsive. This is very inconvenient to attempt to use a computer like this. Having read most of this thread is seems alot like the same issue everyone has been having, and am I understanding correctly that there is no fix for it?

Here are my specs:

2x7200.11 500GB's in RAID0 on an Intel Matrix Raid Controller; I'm not sure if it is 8/9 or 10R, but the driver version is 8.0.0.139. Running Vista 64 Bit.

My 1st hard drive in the array has been reported as degraded now for quite some time;I heard of some users 'rebuilding' their array, I am unsure if I should do that however, as I am not sure what effect that will have on my data. 

Should I rebuild my drive and see if that stops the issues somehow? Or maybe upgrading to 8.8 would be benificial? I would love to know what steps I could take to prvent this from happening to my drives, but I also need to keep my data on the hard drives! Maybe I should be looking at some Samsung F3s...

Any help is greatly appreciated!

Thanks,

Nick

[mooner]

Hi Nik-

Based on the problems I've had (and others in this thread) rebuilding your drive will likely not improve matters...

It seemed that some people were having better luck upgrading (and in some cases downgrading) Intel Matrix firmware...

But you should call Seagate and open a ticket -- your use scenario (what they call Desktop RAID) is actually a usage scenario that they admit is a suitable use for the 7200.11 drives. For some reason, if you're running RAID 5 or more than 2 drives, suddenly Seagate thinks that is not a suitable use for these drives. But in your case, you are using the drives exactly in the manner that they say is suitable.

Word of warning -- if they want to RMA your drive, they'll probably send you a replacement that has the same problems.

It does seem that some folks have had better luck with drives running SD1A/SD1B firmware. Not sure what you have on your drives, but CC1H and CC1J were useless when I put them on an LSI JBOD controller.

Good luck, but we've been waiting for a reasonable response from Seagate for nearly a year...

-E

[Undermoose]

Nick, it's pretty clear that something has changed as the problem has become worse.

If it's the same drive every time then get it replaced.

While there are confirmed reports obviously regarding this generation of disk having issues, and obviously these issues are persistent, in an environment where all things being equal it spontaneously takes a turn for the worse... One must also pursue the normal troubleshooting path.

This isn't the first generation of drives to drop out of raid arrays,

I just experienced my first drop out for a 6 disk raid 10 of the 7200.11 1.5tb CC1H and my first thoughts are my recent northbridge and memory voltage tweeks, but I'm watching this like a hawk now. If the problem persists I'll start rma'ing disks. 6 weeks and no issues though so I'm thinking it's good and was my environment changes.

On a side note, I'm POSITIVE the new intel matrix storage manager is involved in some of these drop outs too. I upgraded my friends computer with the 8.9.1023 and his 3 disk raid started loosing a disk (different one each time), and his array was fine for years.

Now this IS the problem in buying questuonable hardware... You can't put the issues to rest...

Message Edited by Undermoose on 10-20-2009 02:56 PM

[prune998]

Hi,

Just to let you know, after spending a lot of time, and money, and still not beeing able to use the 8x1.5Tb drives,  I went to the shop and bought the new 2Tb Hitachi drives... and they are working like a charme with even a throughoutput increase.

Conclusion is that Seagate s**x, is aware of a problem and just don't care and I will NEVER buy a seagate (nor my company will) EVER AGAIN.

[Edited in compliance of the community rules and regulations.]

Message Edited by BradC on 10-24-2009 10:38 PM

[PaloAltoDan]

Just another voice here -

I've got 5 1.5 TB 7200.11 drives part number ST31500341AS in a a desktop raid 6 array using a brand new 3ware 9690SA controller.  Running under Windows 7 64 bit as a large (4TB) volume which I am using to store video files.

I'm seeing similar problems to others here - the drives randomly drop-out from the array. The controller reports errors like "drive removed" or "command timed out."  A reboot of the machine fixes the problem - as long as no writes were going on at the time the array comes back in normal status.  I can reproduce the problem by running a "verify" operation on the array from the controller - about 15 minutes after the verify starts, one or more of the drives will reliably drop out. 

3Ware support had me try turning off NCQ, which had no impact. They also wanted to make sure there were no power supply issues - there are not, I've got watts to spare. 

This problems seems very reproducable to me - don't understand why Seagate doesn't appear to be on top of it.  I've got close to $1000 of their drives that are pretty much useless.

Message Edited by PaloAltoDan on 10-27-2009 09:02 PM

[fzabkar]

The following thread may provide some insight. See the bad block maps generated by HD Tune after a command timeout.

STM31000340AS disappears from computer:

http://forums.seagate.com/stx/board/message?board.id=ata_drives&message.id=16063#M16063

[PaloAltoDan]

Dear Daniel,
 
Thank you for contacting Seagate.
 
I unfortunately am not able to troubleshoot the drives you have in a Raid environment, this is due to the fact that you are using a 3rd party Raid controller that could itself be causing the issues.
 
I can however recommend to you to test the drives individually with Seatools, this will tell us if the drives themselves are working as they should.  If the drives work outside of the raid environment than something (most likely work load) is going to be causing these drives to fail.  The drives you have are not designed for a heavy raid type workload rather a more consumer level environment.
 
I have seen this issue many times and I am sure it is on many of forum postings however I do not rely on these for my troubleshooting information.
 
I can only suggest that if you are trying to build machines that are going to be utilizing raid that you would look into our more enterprise level drives that are designed and have specific firmware and hardware perks to accommodate for a raid environment.  The drives I am speaking of are the Barracuda ES.2 drives.
 
I do apologize but there is not much more information I can help on this issue.  If you would like to speak to someone about this by phone, feel free to call in and talk to a tech support agent.
Best Regards,
 
Andrew A.
Seagate Technical Specialist