MCS Server SATA Raid1 won't rebuild after drive fail

carlnewton · ‎03-23-2010

Guys,

I have an MCS server with the adaptec hostraid controller on it, the cold-swap serial ATA one. A drive recently failed, so the machine was shut down, a new drive inserted (of the exact same size and model number as the previous) and the machine was booted up.

When looking at the hostraid configuration utility in windows it shows the first raid array as "Degraded" and does not report the second, newly inserted physical drive as being part of this array. It detects the second drive just fine, but does not automatically add it to the array as all of the documentation suggests.

I've had a look through the windows based configuration and the configuration program you can get into at system POST but neither of these seem to have any kind of option to introduce the newly inserted drive into the array.

Can anyone please make any suggestions to me so i can have my redundancy back?! As i say, its an Adaptec SATA controller running Raid1 in a 7825 (only two hard drives, 1u server) On CCM 4.1.3.

Thanks in advance

allan.thomas · ‎03-24-2010

Hi Carl,

As you mentioned these SATA drives are 'cold-swap', and therefore after you replace the failed drive the Array should automatically re-sync. If this is not the case the raid is definately configured for RAID1+0, as you would expect from the initial hardware detection installation, the fact that this fails suggest a problem with possible SATA firmware or BIOS?

I would verify that the BIOS has SATA RAID enabled, take a look at the following post which gives a good explanation of manually checking and configuring SATA:-

https://supportforums.cisco.com/message/330644#330644

As the version is CUCM 4.1.3, I suspect that the OS has not been regularly been updated? If this is the case I would check for known caveats as OS releases can contain BIOS updates, or simularly escalate to TAC, unfortunately I have not found anything conclusive regarding your symptoms.

Hope this helps

Allan.

David Hailey · ‎03-24-2010

Carl,

I have a question about the reboot/rebuild process. This may seem trivial but I have seen some issues with this in the past so that's the nature of the question. When your drive failed and you had to cold swap, did you do this:

Shut down server. Fully insert new drive into slot. Reboot with both disks inserted.

OR did you do this:

Shut down server. Reboot with single good disk. Verify health of server. Fully insert new drive into slot.

I've always followed model number two and believe this to be best practice and also the proper way to rebuild the RAID array after a drive has failed. In the past, even if you had hot-swappable disks - Cisco waivered on support for anything other than cold-swapping so the details of your rebuild process were always worth noting in case it came up during a support call.

If you went the first route, I would attempt to rebuild again but using the second approach. It may not be successful after the fact, but it's worth a shot IMO.

Hailey

Please rate helpful posts!