Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

I am working on a machine that has Intel Matrix Storage Manager on it. It says it needs to rebuild the RAID1 array, but it seems to get to about 35% and then starts the rebuild over again at 0%. That's been going on for a while now. What does this indicate?

This machine was in disarray when I got to it. It may have been cycling like this for a long time, but no one noticed.

This is Intel Matrix Storage Manager 8.9.0.1023 (on on the windows console).

This is an older version of Intel Matrix Storage Manager. I could update it if this might be a software issue. Or, if I'm not mistaken, Matrix has been replaced with Intel Rapid Storage Technology, and moving to that would probably be better?

Else does this kind of cycling indicate that a disk is failing, or that they are incompatible? These are WD drives, WD3200AAKX-0. OS is Windows 2003 SP2.

share|improve this question
    
Before you do anything else, backup the server! This is an unstable condition and could degrade to failure and data loss at any time. –  David Schwartz Apr 30 '12 at 21:50

1 Answer 1

up vote 1 down vote accepted

This is Intel Matrix Storage Manager 8.9.0.1023 (on on the windows console).

This is an older version of Intel Matrix Storage Manager. I could update it if this might be a software issue.

I would be immediately suspicious of the WD Caviar Blue drives like the ones you have. They do not have TLER enabled so are not optimal for RAID arrays. Yes, this might be your problem. TLER disabled drives will appear to drop out of arrays under certain circumstances. You need enterprise drives that have ERC capabilities that support being placed on a RAID controller.

Having said the above, MSM 8 had known issues with random drives being marked as having failed. Indulge me in a bit of hand-waving, but it's almost always advisable to keep current with software, especially drivers and management tools that interface directly with hardware. Given the known issues with MSM and volume failuers, yes, upgrade the product. Yes, use the latest version of Rapid Storage Technology that your chipset supports.

Update your storage drivers and management software. If the trouble persists, then you can focus on the hard drives themselves. If it turns out to be the TLER circuitry, you have no option but to purchase proper drives for placing in a RAID array. Modern WD hard drives do not even have TLER enablement in the firmware thus firmware hacks that previously worked on early generation Caviar drives no longer work.

share|improve this answer
    
That is very helpful, thank you. Am I correct in thinking that there is no straightforward path to migrating the intel matrix software to the more recent version (Intel RST)? I will create a new post re that...if you would respond to that it'd be very appreciated. –  meridian Apr 30 '12 at 21:06
    
I updated to irst 10.8.0.1068 (iata_enu_10.8.0.1003) It did not help with the rebuild issue. In fact, before and after the irst change, the server was just out of it, almost unable to respond to commands and had at least one BSOD. Task Manager did not show the cpu being pegged, but the hdd light was on non-stop. I needed it to work so I have removed one drive and it has returned to normal, though the RAID1 is degraded of course. I may need to replace the drives as you suggest, but I do see others have used them for raid (just checking newegg). –  meridian May 1 '12 at 14:11
    
One thing that I don't understand is that despite the update to irst, the boot time RAID screen remains as it was, "Intel Matrix Storage Manager option ROM v 8.0.0.1038". I expected that to change to something else. My workstation has v 8.9.0.1023. Why would the irst update not impact the boot RAID? The server's boot RAID is older than my workstation's. Does one update the boot raid separately? Would it possibly fix this disk contention issue, if that's what it is, with a ROM update? –  meridian May 1 '12 at 14:11
    
One other thing, NOW when the two disks are installed, the machine takes forever to boot and shut down. The super slowdown started when I noticed at boot time that the raid array said it needed to be rebuilt, and then installed the windows level Intel RAID software that came with the mobo (matrix). As I mentioned before, without 2 disks, it boots at a normal speed. –  meridian May 1 '12 at 14:28
    
It's almost certainly a problem with the TLER circuitry in the consumer grade drives. It's not actually a problem - they're doing what they were intended to do. They just weren't intended to be placed in a RAID set. =| –  Wesley May 1 '12 at 17:46

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.