Sign in with
Sign up | Sign in
Your question

Raid 1 first time critical notice (help)

Last response: in Storage
Share
April 15, 2012 12:53:10 PM

Hi all,

first some history, i burn through an HD about once a year for the past 3-4 years, no idea why and thats an interesting question on its own but for another time. About a year ago my HD failed as usual and i replaced it under warranty, when i got the new HD i bought another identical model and setup a raid 1 array.

Drives: 2x 1TB WD Caviar Green (WDC WD10EARS-00Y5B1)
MB: Gigabyte GA-MA770T-UD3P

Setup on a bios raid (Ctrl+F on startup)

2 days ago when i started my computer (after a normal shutdown) it told that one of the drives is crytical.
After windows loaded i started googling downloaded RaidXpert and started a rebuild (and googled some more) which is now at 94% (after ~50 hours).
Both drives are listed in RaidXpert as functional and SMART healthy.

Was this crytical just a glitch ? it got me really scared and im not entirely sure what to do.
Also "Western Digital Data Lifeguard Diagnostic" cant run while the array is set up, can i break the array, test each drive seperatly and reconstruct the array ? if so how would you suggest to go about this procedure.
a b G Storage
April 15, 2012 2:24:39 PM

unplug the 1 drive and you should be able to run a surface scan to verify its integrity. and did you try to run the WDDLD while it was rebuilding?
April 15, 2012 3:13:01 PM

ill try unpgugging without breaking the raid array after the rebuild.

no i tried to run wddld b4 the reebuild started.

update: at 98%. got this warning at 95%: Task 30 timeout on disk (Port Number 1,Target ID 1) at LBA 0x06dff6e80 (Length 0x80). Any thoughts on what it means ?
Related resources
April 15, 2012 3:37:23 PM

The first thing i would do is stop using that on board software raid, and get a really good mechanical raid card from 3m or some other good manuf. They are leaps and bounds faster, and much more reliable. If you're data is important to you, that is my suggestion. Any raid setup with only 2 drives, that takes 50 hours to re-build, is simply not even remotely adequate in my opinion. Yes, I know, a good raid card can cost $300 plus dollars. But, with a good raid drive enclosure, and good drives, they are un beatable interms of speed, reliability and performance. I've been using a similar set-up wiith 5 drives, for several years now. To make a long story short, software raid is not the way to go.........

As for being able to remove a drive, and check the remaining one, etc.....it completely depends on what mbod you are using, and whether or not it is hot swappable, etc. Some boards would be ok with that. Some, amy not. There are risks with this. by the way i've been using the same drives in my raid for three years with no failures..... And, if speed is important to you, i wouldn't use the WD green drives. Tey are energy efficient models, and are not meant to be perfoamnce drives. My choice, if I weere to buy WD, would be the black ones.

I hope you are able to get all your data back, and that it all re-builds ok for you..... best of luck!
April 15, 2012 4:53:42 PM

I am not too concerned with speed/rebuild time, the raid array served its purpose perfectly and that is that my computer doesn't go bye bye when the drive (single - no raid) fails. It is still running and the rebuild is going on in the background so i am happy with my array in that respect.
As your array is 5 drives im assuming that the goal of our arrays are different and thats fine :) 

as for the hotswap ill check, tnx.

E/UPDATE: the rebuild is done and the logical drive is functional (ie not crytical), i gess i panicked for no reason but i wanted to check them for bad sectors, as im unsure if it supports hotswap and such i was wondering if a media patrol could replace my check ?.

Also how long should it take and how often should i schedule it ? currently its not schedules.
April 15, 2012 11:53:04 PM

There was some FUD around using WD drive in raid arrays. Google 'time limited error recovery' Not sure if this was real, or just a trick to get people to buy the very nice, fast and expensive 'raid edition' RE drives from WD. I have a WD black in raid 0 with a seagate drive, no problems for about a year now.

I bought 2 WD green 1TB drives for some crazy reason 2 years ago. They are both dead now. The Black drives I bought were only a little more expensive, performed way better and are still alive. Your experience with green WD 1TB drives mirrors mine. They may have had a bad batch/design.

re "Task 30 timeout" sounds like the drive was not responsive for longer than the controller would tolerate, but that retrying worked. There used to be a bunch of errors that cause the drive logic to reset that could be long running (issue IO, lose response, wait to timeout, issue reset, wait for reset to complete, re-issue IO) plus there used to be thermal recalibrates where after the drive heated up teh head wouldn't go tot he right spot, etc. Not sure modern drive has these problems. I'm out of date.
!