Failed RAID 1 - Intel Rapid Storage Enterprise SCU, No recover option

fyrye

Honorable
Jul 1, 2013
136
0
10,760
I have a SuperMicro server with the intel rapid storage enterprise SCU using a RAID 1 (mirrored).
One of the drives failed but I am unable to determine which.
The status of the drives is "bootable - no".
In the original configuration the list of drives is:
SATA 0 - Error Occurred(0)
SATA 1 - Offline Member

Removing either drive to boot from a single drive and swapping ports results in Error Occurred(0) still as bootable - no.

I attempt to recover the volume like I used to in the Intel Matrix Storage Technology setup in the BIOS, but there is no prompt to recover, when entering the raid configuration settings for the failed volume.
This prevents the ability to have redundancy that I was expecting from the mirrored drives, if one fails, the other should run in it's place notifying me of the degraded status like the Matrix Storage technology would do. But for some reason the Rapid storage SCU fails to say degraded at all.

I even tried two different drives and they are detected fine by the rapid storage SCU to see if it was a faulty controller.

I decided to test the drives in an older server that has the older matrix storage technology instead of rapid storage scu and it prompts me to recover the failed volume.

I don't believe it would be safe to attempt a recovery this way. Any suggestions on how to recover the failed mirrored volume?

I went to call intel for support but they are closed on the weekends.
 
Solution
...

fyrye

Honorable
Jul 1, 2013
136
0
10,760
I took both drives and mounted them in a LiveCD environment to determine which drive was faulty.
One of the drives would fail to read while attempting to access it, the other performed fine.
Attempted a quick diagnostic using a few different applications, all of which failed the defective drive.

I took the functional drive and used the previous server's Intel Matrix Storage Technology software.
Upon entering the Raid configuration it prompted me to recover the array volume.
I accepted and it simply marked the drive as degraded and set bootable - yes.
Put the degraded drive into the original server, Rapid Storage SCU detected it as degraded as well and booted into windows fine.

Awaiting replacement drive to rebuild the mirror.
 

TyrOd

Honorable
Aug 16, 2013
527
0
11,160


This is the standard way of doing it. I don't see anything wrong with what you did here.

However, always remember that RAID is not backup, so it won't protect you from data loss like a proper independent backup scheme.

The purpose of RAID is availability so if you're having issues with this setup I would make sure that you can rebuild properly in the future as to minimize the kind of downtime you've experienced.
 
Solution

fyrye

Honorable
Jul 1, 2013
136
0
10,760


Thank you TyrOd.
My main concerns was bricking the drive by using a different version of software to recover the array, prompting two RMA's instead of one.
The issue I experienced in this instance was the raid 1 volume should have degraded instead of "errored", and the Rapid Storage SCU ROM did not provide any means to recover from the failure.
Instead I was forced to use the secondary server with a much older ROM raid controller version to recover the functional disk from the issue, since it properly recognized the error and prompted me with a recover option.
I apologize for not being more clear as to the initial issue in answering my original post.

As for the expected raid's configuration purposes you are correct and I utilize an iSCSI SAN to store data and perform onsite backups, as well as offsite cloud storage to do incremental backups with file delta redundancy and encryption.
I am also setting up a failover server to work with the SAN data in order to provide additional hardware redundancy in case any similar issues occur again.
The raid volume that failed only contains the OS and I was expecting the hardware redundancy of raid 1 to compensate for the single disk failing. Data was never the issue.