Question : Novell error code after hardware failure

Hi,

I had a DL360 with a SmartArray 5i card running NetWare 6.0Sp5 suffer a hard drive failure.  The drives are mirrored and the mirror broke leaving an error on the console indicating that a drive left the array.  So, I had a new drive sent to me.  It's a hot swappable drive so I popped it in.  Here is the error:

 1-18-2006  10:46:32 am:    CPQSHD-2.2-0
   Severity = 3
   CPQSHD: The HP Smart Array 5i Slot 0 ID 0 LUN 1 disk device
failed to initialize.  This device will be deactivated.

The error kept repeating at an alarming rate and I pulled the drive.  The server kept functioning normally.  Now, there is an orange light on in the server indicating an error and the drive is still pulled.  Obviously, I'm too chicken (or wise) to put the drive back in and possibly have it rebuild the array again but this time, fail completely and lose the logical drive.

The HP tools are not loaded on the server nor is Insight Manager.

Three questions:
If I boot the server up with a raid utility, will I hurt the logical drive?  
Can I rebuild the array (put the mirror back in place) without the OS on (booted from disk) and without losing the logical drive?  
If I find I have to replace the controller, will I lose the logical drive and the data?

Thanks very much for your help.

Answer : Novell error code after hardware failure

Judging from the error message, you have a hardware-based array. NetWare is only aware of the RAID array because the SmartArray 5i driver tells it. The server is functioning normally because the RAID array is being done in hardware, and as long as you don't loise another drive (assuming its RAID 5), you'll be fine.

I would NOT put the drive back into the server. If the RAID controller says its failed, deactivated it and kicked it out of the array, then its probably bad. Putting it back in will, at best, do nothing. At worst, if another drive fails while the RAID controller is trying (without success) to rebuild the drive, you'll be hosed.

While it is true that back in the days of the Compaq SmartSCSI-1 Array controllers the loss of the RAID controller would also lose the array, the logical drive info nowadays is stored on the drives as well as the array controller. As long as you replace the 5i with another 5i, and the firmware versions are compatible, you'd be fine. But I don't think, given that the server is running fine, that the controller board needs to be replaced. The drive needs to be replaced. The controller board is probably working just fine.

Get a replacement drive and stick it into the same slot as the failed drive. The array controller should rebuild the array in the background, NetWare will never know the difference. The only way you'll screw up your array is if you move drives to different slots. Don't do that - you WILL lose the array.