SA641 on ML350G3, Waiting for rebuild, what to do?

[Joost van der Valk_1]

I have a ML350G3 fitted with a Smart Array 641 controller.
There are four 36GB drives installed as a single Raid 5 logical drive with one drive as online spare (id3).

Appearently something went wrong with drive id2 2 weeks ago, because the online spare was activated. (hp management logs)

I didn't notice this, because it's a remote server, and I rebooted the machine from remote.
It didn't come up. When I drove over, it showed the F1 or F2 question on the console, asking me to rebuild or not.

I chose not to rebuild.

Strange thing is that after booting, the ACU now says all 4 drivers are OK, and the status of the logical drive is Waiting for rebuild.

Drive id's 0, 1 and 3 have the cilinder light burning, drive id 2 has only the arrow light burning.

I think, but I'm not sure, that the controller now thinks drive id2 is OK again, and wants to rebuild back to drive id2 and switch the online spare (id3) off.

I ordered a replacement for drive 2, don't trust is anymore, and don't know what to do now.

Reboot first and let the controller rebuild first to id2 or first replace disk id2 and then let it rebuild (automatically?).

Any tips?

Thanks in advance,
Joost van der Valk

[amhakassa · Accepted solution]

Hi Joost,

If you have already a new replacement drive for the failed drive in ID 2
1. Power down the server
2. Pop the new drive in place of the failed drive in ID 2 and reboot the server
3. When you got the F1 or F2 option on POST choose the rebuild option.

This way the drive in ID 2 will be rebuilt and your onlinespare remains as an online spare.


Regards
Amha Kassa

[e4services]

OK, there is something that is not clear. You said "four 36GB drives installed as a single Raid 5 logical drive with one drive as online spare"
That is 5 drives. You only mention 4 id0-3. So I guess then it was 3-RAID5 (0,1,2) 1-Online Spare(3). (2) was replaced with (3). (2) went offline.
My guess, if this is true, that (2) was a predictive failure and was replaced.
You interrupted the rebuild using (3) and it wants to continue.
(2) did not give the predictive failure result when rebooted (this can happen for drives give false results to the controller)

Suggestions:
Allow the rebuild
Replace drive (2) anyway. It's your data.

[Joost van der Valk_1]

To Amha Kassa:
Thanks very much for the advice. I did exactly as you stated and it worked out very well. No problem at all. The new disk is active now and the online spare became spare again.

To e4services:
You are right, it is a 3-RAID5 (id 0,1,2), 1-Online Spare(id3) setup. Id2 was replaced with id3. Id2 went offline 2 weeks ago. Thanks for your reply.

Regards,
Joost van der Valk