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

[ardent]

Hi designer.  I am the one who posted the Intel Storage Matrix 8.9 problems and that I found some in the intel forums reverted to 8.8 and their problems went away.  

I almost didn't post my findings here but I wanted to share what I had found.  I even had to wait 2 days for seagate to reset my account as it had been disabled after years of inactivty. :-)

Anyway, please continue to post your updates, my ICH10R onboard the Asus P6T is running still without any more drives falling out.  *knock on wood* I hope it continues, but, like i said in the other post, it ran fine since February before I reloaded OS (for other reasons) and grabbed the latest drivers (Intel Storage Matrix 8.9).

Good luck

Edit: here is the Intel thread: http://communities.intel.com/thread/5036?tstart=0

Message Edited by ardent on 08-05-2009 07:11 PM

[HughR]

ardent: thanks very much for posting.  This looks like the first solid breakthrough.

The Intel drive problem appears to hit more than Seagate 7200.11 drives.  So it doesn't look to be a Seagate problem.

It is too bad that Intel hasn't acknowledged the problem.  Sound familiar?

Unfortunately several of the problem reports in this thread and others don't involve that driver.  So Seagate isn't off the hook.

I've updated message 1 in this thread to ask for OS and driver information.  I don't know why I left that out before.

I've updated message 3 in this thread to mention the Intel Matrix drive problem.

It would be great if we could tell which reports might be addressed by a driver downgrade.

Message Edited by HughR on 2009-08-06 10:58 AM

[KerryRJ]

Since my previous post I have tried Intel Storage Manager 8.9 and still have a random drive that drops out of the array and either automatically rebuilds or I have to pull the drive and reinsert it to get it to rebuild.

I have also replaced 4 drives at a customer's site with equivalent 7200.12 and have not had any drives failing or dropping out of the array but it's still too early to tell.

At this point I am not convinced that the Intel Storage Manager 8.8/8.9 is the problem.

Guess time will show if the 7200.12s work or don't.

[HughR]

---

KerryRJ wrote:

Since my previous post I have tried Intel Storage Manager 8.9 and still have a random drive that drops out of the array and either automatically rebuilds or I have to pull the drive and reinsert it to get it to rebuild.

---

Do you mean that you've tried 8.8?  Because 8.9 is the one that seems to cause problems.

It is probably worthwhile mentioning your OS.  I don't know if the same drivers are used in all versions of MS Windows.

Message Edited by HughR on 2009-08-06 11:01 AM

[KerryRJ]

Woops didn't notice that I hadn't included the OS.

Windows 2008 SP2.

Problem was also occurring under Windows 2008 SP1

Yes have tried both Intel Storage Manager 8.8 and 8.9 and both have had drives drop out although 8.9 seems to have fewer drives drop out of the array.

Just found that there is a new Intel Storage Manager 8.9.1023 dated 2009/07/17.

Have upgraded to the new version from 8.9.1006.

Message Edited by KerryRJ on 08-06-2009 09:10 AM

[prune998]

Guys,

I'm really sorry, but I really don't think the "software" part will solve your issue. At least, it will make the problem aprear later.

For my part, I'm on a Mac, OS X 10.5.8, with drives connected to a Highpoint 4322 RAID card, with intel IOP processor.

The only special thing to know about this chip is that it does not support SSC functionality.

I had a look at the new 2Tb western digital spedial RAID drive today, and some guys seems to say it also does not work in RAID arrays or with some chips. This would be due to the "green feature" that slow down the disk when he think it can. This seems to fool the controler. This is even more clear in a RAID array... Would the seagate drive have the same kind of problem  ?

[ardent]

Re: Intel ICH10R.  This 'software' also bundles drivers for the raid card, so it likely is more than just 'software.'  Sorry for not being so detailed.

Still running great with no problems on 8.8!  

Edit: 8/8/09 Still no problems...

Message Edited by ardent on 08-08-2009 04:23 AM

[HughR]

AlanM has not seen a pattern of 7200.11 drives going offline.  I'm trying to help gather the data so we can see if he is right.

I've started a new thread to report instances of non-RAID disks going offline.  Please add suitable reports to this  7200.11 non-RAID drives going offline: unified thread

I've spent several hours(!) adding the tag "freeze" to all reports in the enormous firmware thread that seem to indicate the drive going offline.  Seagate Barracuda 7200.11 Firmware Issues

It would be great if all people reporting problems in this thread tell us whether or not they were using the Intel Storage Matrix 8.9 driver since that driver appears to cause similar symptoms.

[mooner]

Hi Hugh-

Ok, in my case, I am not using an Intel RAID chipset.

My most troublesome controller is a Highpoint 2322 PCIe to SAS/SATA RAID controller. I am also using LSI controllers with mixed results.

But, as I have noted before -- on the Highpoint 2322, I am using the Seagate ST31500341AS drives in "Legacy" JBOD mode. My drives do NOT suffer from going off-line, so I cannot respond to your new threads.

My drives are simply ridiculously slow. File transfers with files of any significant size (more than a few MB) will usually timeout and then the OS reports errors on the drive. Sometimes the drives will no longer be accessible, sometimes they still work. Generally, reading is much better than writing to these drives, but even reading is generally slower than expected and not entirely without similar problems. Read speeds typicall top off around 40-45 MB/sec (when it is working well) and Write speeds always languish around 6-9 MB/sec and usually go to zero if writing a big enough file.

Interestingly, as I mentioned earlier in this thread, I purchased an ST32000542AS (Baracuda LP series) 2TB drive to test and see if it has the same problems. I connected it to the same Highpoint 2322 controller (since that one is the source of most of my problems). So far, it seems to WORK FINE. My 2TB drive shipped with firmware CC34 and I have not changed that. The WRITE rate is > 70MB/sec -- that's a 5900 RPM drive moving data 1.6x faster than the READ rate on my 7200 RPM 1.5TB drives on the best day. I've only been testing for a few hours, but so far the 2TB drive is taking my 20-50 GB files with NO write problems at all. Exact same PC, RAID contoller, SAS enclosure, cables, OS, etc.

So, as I have stated on many occassions, I believe my problems all stem from the 7200.11 series drives...

I just have no idea why it is so difficult for Seagate (or Alan for that matter) to understand this...

Message Edited by mooner on 08-08-2009 10:18 PM

[designer]

Well, it's been running for a couple days now on 8.8, and so far no drives have dropped out   The only change was going to 8.8 drivers.

I was (somehow) running 8.9 drivers, and 8.0 Intel Management software, though I doubt that had anything to do with the problems.  I'm now running 8.8 drivers/manager.  

I also noticed the overall speed about doubled.  With the 8.5 drivers, I was only seeing reads of around 100mb/s.  Now it's around 250ish.  

I would not be surprised if something in the drives is running at a borderline spec, and something in the 8.9 drivers pushes it over the edge.  That would account for the problem with those drivers, and on systems that don't even have intel chipsets (they could be doing the same thing the 8.9 drivers are). 

When things are running as fast as they are, it doesn't take much before in-spec becomes out-of-spec.  :shrug: