We have a DL380 had system board failure. Specifically, the disk controller. Took a couple trips, but they got the machine back up and running. However, there is a problem with a RAID 1+0 array on the server.
The file system on this Raid is up and apparently fine. Running RedHat.
The problem is that the array thinks it has 3 failed disks, and the controller thinks there are 3 spare disks.
array B (SAS, Unused Space: 0 MB)
logicaldrive 2 (1.6 TB, RAID 1+0, Interim Recovery Mode)
physicaldrive 1:5 (box 1:bay 5, SAS, 0 MB, Failed)
physicaldrive 1:7 (box 1:bay 7, SAS, 0 MB, Failed)
physicaldrive 1:8 (box 1:bay 8, SAS, 0 MB, Failed)
physicaldrive 2C:2:7 (port 2C:box 2:bay 7, SAS, 300 GB, OK)
physicaldrive 2C:2:8 (port 2C:box 2:bay 8, SAS, 300 GB, OK)
physicaldrive 3C:2:2 (port 3C:box 2:bay 2, SAS, 300 GB, OK)
physicaldrive 3C:2:3 (port 3C:box 2:bay 3, SAS, 300 GB, OK)
physicaldrive 3C:2:4 (port 3C:box 2:bay 4, SAS, 300 GB, OK)
physicaldrive 4C:1:1 (port 4C:box 1:bay 1, SAS, 300 GB, OK)
physicaldrive 4C:1:2 (port 4C:box 1:bay 2, SAS, 300 GB, OK)
physicaldrive 4C:1:3 (port 4C:box 1:bay 3, SAS, 300 GB, OK)
physicaldrive 4C:1:4 (port 4C:box 1:bay 4, SAS, 300 GB, OK)
physicaldrive 5C:1:6 (port 5C:box 1:bay 6, SAS, 300 GB, OK, active spare)
unassigned
physicaldrive 5C:1:5 (port 5C:box 1:bay 5, SAS, 300 GB, OK)
physicaldrive 5C:1:7 (port 5C:box 1:bay 7, SAS, 300 GB, OK)
physicaldrive 5C:1:8 (port 5C:box 1:bay 8, SAS, 300 GB, OK)
Doesn't take long to see that there is a house keeping issue here.
We tried reseating one of the drives to see if that would get the controller to figure it out, but that didn't happen.
My next thought is that we could go in hpacucli and remove the "failed" disks and add in the spares.
- Ctrl slot=0 array B remove drives=1:5,1:7,1:8
- Ctrl slot=0 array B add spares=5C:1:5,5C:1:7,5C:1:8
Then, I'm not sure what command to use to get the spares to take over and be full members of the array and not just spares.
Is there any danger to this? If we have to, we can delete the disk and array, and rebuild them, but we already had a week long outage due to the failure, and would rather keep the machine up.
Any help would be appreciated?
thanks,.
Paul.
