Raid 5 Constantly degrading...

[creationsof12]

Hello,

I have searched around for a while and cannot seem to find an answer for my particular situation.

I have a 3-hard drive RAID 5 setup that is constantly saying it is degraded. I have done multiple tests on each of the drives and they are all working perfectly. Just in case though, I replaced all drives with new Samsung drives, re-installed Windows and still... same problem. Also: Each time it gets degraded it says it is a different drive.

The one way that I have found that works temporarily is: shut down and unplug the computer, unplug the "degraded" drive, turn computer back on and open Intel Matrix Storage Console, shut the computer down again, plug the "failed" drive back in, turn computer on, open Intel Matrix Storage Console and then it says the drive is working properly and begins the rebuilding process (which takes around 12 hours).

I am not absolutely sure but I think it may be the motherboard (hopefully not) or the Intel Matrix Storage Console software. I will be in the middle of doing something (doesn't seem to matter what I am doing) and the computer will just lock up and say the volume is degraded (sometimes it says it while I am still on the computer; other times it says it after I restart).

SPECS:
-------------------------------------------------------------------------------------------
Motherboard: TYAN S7002G2NR-LE Dual LGA 1366 Intel 5500 Tylersburg SSI CEB Dual Intel Xeon 5500 Series Server Motherboard.

RAM: Kingston 24GB (6 x 4GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10600) Desktop Memory Model KVR1333D3N9K3/12G

Hard Drives: 3 Samsung Spinpoint F4 HD322GJ/U 320GB 7200 RPM 16MB Cache SATA 3.0Gb/s 3.5" Internal Hard Drives setup as RAID 5 (as I said: It doesn't matter which hard drives I try, I get the same results).

Graphics card: nVidia GeForce GTX 285 1GB.

processors: Dual Xeon 5500 series 1366 quad-core hyper-threaded Intel processors.
-------------------------------------------------------------------------------------------

Any comments or suggestions would be much appreciated!

 

[GhislainG]

Have you tested with drives designed for RAID use? Same problem?

 

[creationsof12]

Thanks for responding.

Yes I have and yes, same problem.

 

[tsnor]

I was looking to see you had WD blacks, they are known for this. The Raid Edition WD's limit the amount of time they spend on error recovery so that the raid controller doesn't drop them. The WD blacks do not allow adjustment of the recovery time.

A quick google of Spinpoint F4 and raid sees no such problem with the F4.

The rest is speculation...

Are you happy with your connectors? cables? temps OK?

Is your Power supply OK, drives all spinning up at the same time used to push PSU enough there were options to stagger startup in the bios. Haven't seen that in a long time. Three disks shouldn't be a problem -- is this a 24x7 system or does it start / stop ? What is the power mgt setting for disks -- do you spin them down after inactivity ? Do you have a good UPS with boost/trim you can use to see if line voltage hits are hurting your system (the array fails every time you vacuum clean or make drinks in the blender)

aside: I came across this "You've turned write-back caching on, right? That will also enable read-ahead caching. " It explained why someone was getting 1/2 the sequential read performance expected form raid5. I would not have guessed that write caching was tied to prefetching read data.... http://forums.storagereview.com/index.php/topic/26670-intel-ich9r-raid-5-3x-1tb-harddrives-problem/page__st__10

Posting to the storage forum might get you better results that the homebuilt forum (or Moderator could move...?)

 

[Outlander_04]

If you are committed to using RAID then my suggestion would be to buy a hardware RAID card and run through that instead of the motherboards RAID

 

[creationsof12]

I had Western Digital drives (not the blacks) before and had read that they were not the greatest for RAID setups (which is why I had thought that was the problem) and ended up replacing them with ones that were meant for RAID setups.

Connectors, cables, etc. are all good (replaced all with spares I had and then put them back).

Power supply was replaced/upgraded... problem continued. I'm not sure about the staggering options but I can look into it. As of now, it appears that all drives are spinning up at the same time.

The machine can pull about 340W. The breaker for this room does sometimes flip off when several things are running at once but I have a good UPS (which remains on and fully charged) that keeps the machine running. I would think that this was causing the RAID degradation except for the fact that I had previously had this machine at another location where the 30Amp breakers (instead of the 20Amp breakers this section has) never flipped off (still had degradation issues) and It still happens when I have everything else in the room shut down (and breaker doesn't flip off).

I would be grateful if a moderator could move this post to the storage forum. I wasn't sure where to post it as I wasn't sure what the problem was.

 

[creationsof12]

If you are committed to using RAID then my suggestion would be to buy a hardware RAID card and run through that instead of the motherboards RAID

I have heard a lot of good things about RAID 5 and would like to continue to use it... but I am seriously thinking about Ghosting my 3-drive RAID 5 to a single disk with a weekly image backup. I'm sure RAID 5 works great for most people but this system just doesn't seem to like it.

I'm thinking more and more that it might be the motherboard. If it is and I want to keep my Windows 7 setup for the new motherboard then I guess I could ghost the RAID to a single drive, setup the new RAID on the new computer and then ghost it from the single drive to the new RAID... right?

 

[GhislainG]

You wrote that you had WD non-Black. Were they RE4 drives? It's the only WD SATA drive designed for RAID use.

 

[creationsof12]

You wrote that you had WD non-Black. Were they RE4 drives? It's the only WD SATA drive designed for RAID use.

The old drives were Western Digital Caviar Blue WD2500AAKS 250GB 7200 RPM 16MB Cache SATA 3.0Gb/s 3.5 drives which really should work in a RAID setup but I had read (after I had purchased them) that they were "not the best for RAID setups." Those should not matter now though because I am using the new Samsung drives that are meant for RAID setups.

 

[tsnor]

What is the power mgt setting for disks -- do you spin them down after inactivity ? If so as a test turn this off.

 

[creationsof12]

What is the power mgt setting for disks -- do you spin them down after inactivity ? If so as a test turn this off.

Ah, it was turned on and set to 20 minutes. I had disabled this a while back and am not sure how it got turned back on. It normally degrades after a couple to a few days after rebuilding process and it just finished rebuilding so it may be a few days before I will be able to notice if this had an effect.

Really hope this will solve the problem and thank you for the suggestion.

In the meantime: Is there anyone else having similar problems? Is there anyone having problems with this motherboard or Intel Matrix Storage Console software? Anything else it could be (just to cover all angles)?


Thank you all for your comments and replies. It is GREATLY appreciated!

 

[creationsof12]

(EDIT) ..... I spoke too soon. It is at it again, raid volume degraded...