Problem with WD20EARS drive in RAID 5

mdutra

n00b
Joined
Feb 14, 2011
Messages
2
I purchased 4 WD20EARS drives on sale to create a more fault tolerant storage system for my home theater (movies, recorded television, music, etc) since my last external drive crashed losing all my data.

Now that I have the drives I read that pretty much of any drive out there the WD20EARS is the one NOT to use in a RAID. However, I have already purchased them. I am running into an issue but I do not see what all the posts say I will see which is the drive falling out of the array. What I see is is the following event in the system event log from time to time:

"Error Source: rr62xx, ID: 9, The device, \\Device\\Scsi\\rr62x1, did not respond within the timeout period"

I would have ignored this except about half the time this error appears in the log strange things happen to the server. I can no longer log on remotely or locally but remotely I can see the drives and access some file, even on the RAID drive. However, others when double clicked simply hang indefinitely. When I do try to log on locally I can type in credentials and then enter, screen then goes blank and stays there until I hard power the machine. When it boots back up there is nothing of note in the event log (as far as I can tell). No errors, warnings, just a few informational about different (seemingly unrelated) services starting or stopping such as web server stopping due to inactivity or Microsoft Software Shadow Copy Provider stopping, etc. The one exception is that within a few hours (or minutes) prior to me noticing the server has stopped responding is the above error with the rr62xx error.

I cannot seem to recreate the error when I want to but it does seem to happen when there is a lot of file processing such as file copies or unzipping. My question is whether or not the drives are the issue here or if there may be a driver issue for the rr622 card. I used it shortly on another machine in the exact same drive config with no issues on a Windows 7 32 bit OS while now it is on a Windows 7 64 bit OS but I also hit the drive a lot more now so I cannot say it is the driver either. I am considering replacing all 4 drives and selling these but I would like to know that the drive are really the problem. Any debugging advice or similar issues out there?

Current server:
Dell T110
8GB non-ecc ram
Highpoint RocketRaid 622 raid card
4 WD20EARS 2TB drives
Windows 7 Pro 64bit
 
Search TLER and you will see what you problem is. Dirves are timing out.....
 
what ^^^ said, the drives take too long to respond to raid controller which causes time outs, eventually those drives will be kicked out from the raid array and the array will fail. Classic TLER problem.
 
Yep, definitely TLER. My advice? Either replace the drives with better behaved units (Hitachi, Samsung, Western Digital REx-GP, Seagate (and in that order too...)), or put your RAID card in JBOD mode and pick a server OS that supports software RAID and/or storage pooling.

OpenFiler, FreeNASand Amahi Home Server are good (and free) candidates.
 
OS X, Linux, and FreeBSD also support software RAID which will preform very well with that setup. Just make sure to setup Samba (SMB) shares so other Windows systems can see your files.
 
I'll just toss this out there (and this is my media server/utorrent box)...

I have 2xWD20EARS.
I am on a GA-MA785GM-US2H (http://www.newegg.com/Product/Product.aspx?Item=N82E16813128394).
Windows Server 2008 Enterprise (not R2).

They're in software raid (RAID1) using the onboard raid controller. I formatted the drives individually within windows before raiding them to hopefully take advantage of the 4k blocks on these new ones.

After that, I used the onboard raid bios to set them up and it worked.

I've not had any errors or warnings in my event logs and it's been running for, hmmm, I'd say about 3 weeks (had to replace the old mobo).

I didn't do anything special really. Hopefully something in there helps you find a solution to your own problem.


Of course, when my evaluation of Server 08 expires, I will be moving to a linux-based OS. Though that's a bit off topic.

TL:DR = My setup worked, here's what I did, hopefully it helps.
 
RAID 1 doesn't have any parity so you probably wont have an issue.
 
RAID 1 doesn't have any parity so you probably wont have an issue.

Yes, these consumer WD drives are safe to use in single-level RAID 0 or RAID 1 configurations. However, their lack of TLER hampers performance and reliability when used in parity RAIDs (e.g. RAID 3, 5 or 6) or in multi-level RAIDs (e.g. RAID 10 or 30).
 
The drives should perform fine in raid5. Search for wdidle3 and set the head parking time to 300seconds. See if it fixes your problem, it should.
 
It relieves me that pretty much everyone agrees it is the TLER issue. I will give wdidle3 a shot. Isn't that going to update the firmware of the HDs as well though? Any potential issue with moving each drive to another machine to perform the update and putting it back into the array? I wouldn't think there would be but I also didn't think certain HDs have capabilities that are locked out by the manufacturer...
 
Back
Top