I had a failed disk in the Sun storagetek 6540
It's a controller with three expansion trays
the failed disk was t1d08 which is the disk 8 in the tray 1
but the crazy thing that it has affected most of the volumes , & virtual disks
now we replaced this disk temporarily with another disk from tray 2 which was an unassigned disk
and nothing happen
I can't even revive the disk , reconstruct the disk or do any operation on the disk except for assigning it as a hot spare
the crazy thing that of course this disk isn't in all the virtual disks because It's well known that disk can't spin many virtual disks
but the more crazy thing that when we look through the failed virtual disk you can see all the disks are optimal and enabled and nothing wrong with them
except for the really already affected virtual disk
more over does any one advice me with revive the disk or reconstructing the failed disk ?
Which ersion of CAM used ?
Which RAID type was configured ?
supportData from array give more information about current state.
It's very strange that fault of one disk can affective more than one virtual disk.
On every DISK at 6540 writen DACstore - some data about configuration volumes etc.
So when disk change location at array it's not cause change role of this disk.
For Disk from tray2 was writen that this disk unassigned for this array. This function undepend of location this disk in array.
For replace failed disk to unassigned disk you can use command:
service -d <deviceid> -c replace -t <current disk> -q <replacement disk>
Edited by: Nik on 12.03.2012 15:03