Adaptec 133U2 - AAA RAID Controller User Manual page 101

As long as the array member is physically functional, you can
reconstruct it with the Reconstruct command. You can reconstruct
the array immediately, or you can schedule the reconstruct to occur
at a later time. Since the reconstruction process may take a long time,
you may want to schedule it for a time when there is less activity on
the system.
Note:
The Reconstruct operation can take up to several
hours for a large array.
Follow these steps to reconstruct an array when a disk fails:
1
Determine which array is in Critical status and which disk in
the array has failed.
It is possible that a RAID 0/1 array in Critical status may have
more than one failed disk. The array can still be reconstructed
without data loss as long as at least one disk of each mirrored
pair is still good.
2
If the array enclosure does not support hot swapping
I/O to the array before you continue. (See Pausing I/O on an
Array on page 7-5.) Skip this step if the array enclosure
supports hot swapping.
3
Remove the failed disk and allow I/O to resume.
4
Insert a good disk of at least the same storage capacity. Be sure
that the SCSI ID of the new disk is different from the SCSI ID of
other installed devices. Or, if an array disk was accidentally
disconnected, reconnect it.
5
Issue a Rescan command to detect the new disk.
6
If you installed a new disk (or disks), make the disk into a
spare by following the directions in Creating Dedicated Spares or
a Spare Pool on page 5-14. Be sure the new spare disk is at least
as large as the smallest array member.
1
Hot swapping support means that the array enclosure electrically isolates the
bad disk's SCSI connector from the SCSI bus while the disk is being
swapped to prevent data corruption. Data can still be transferred to and
from the remaining good disks while the bad drive is replaced.
Performing Array, Spare, and Disk Operations
1
, pause
7-3