Table of Contents
Advertisement
To minimize the likelihood of fatal system errors, take these precautions when removing failed
disks:
Do not remove a degraded disk if another disk in the array is offline (the Online/Activity LED
is off). In this situation, no other disk in the array can be removed without data loss. The
following cases are exceptions:
When RAID 1+0 is used, disks are mirrored in pairs. Several disks can be in a failed
condition simultaneously (and they can all be replaced simultaneously) without data loss,
as long as no two failed disks belong to the same mirrored pair.
When RAID 6 (ADG) is used, two disks can fail simultaneously (and be replaced
simultaneously) without data loss.
If the offline disk is a spare, the degraded disk can be replaced.
Do not remove a second disk from an array until the first failed or missing disk is replaced
and the rebuild process is complete. (The rebuild is complete when the Online/Activity LED
on the front of the drive stops flashing.) The following cases are exceptions:
In RAID 1+0 configurations, any disks that are not mirrored to other removed or failed
disks can be simultaneously replaced offline without data loss.
In RAID 50 configurations, disks are arranged in parity groups. You can replace several
disks simultaneously, if the disks belong to different parity groups. Do not replace more
than one disk at a time from the same parity group.
In RAID 6 (ADG) configurations, any two disks in the array can be replaced simultaneously.
In RAID 60 configurations, disks are arranged in parity groups. You can replace several
disks simultaneously, if no more than two of the disks being replaced belong to the same
parity group. Do not replace more than two disks at a time from the same parity group.
Replacement disks must have a capacity no less than that of the smallest disk in the array.
Disks with insufficient capacity are failed immediately by the controller, before data recovery
begins.
Automatic data recovery (rebuild)
When a physical disk is replaced, the controller gathers fault tolerance data from the remaining
disks in the array. This data is then used to rebuild the missing data from the failed disk onto the
replacement disk.
The rebuild operation takes several hours, even if the system is not busy while the rebuild is in
progress. System performance and fault tolerance are affected until the rebuild finishes. Therefore,
replace disks during low activity periods when possible. In addition, be sure that all logical drives
on the same array as the disk being replaced have a current, valid backup.
If more than one disk is removed at a time, the fault tolerance data is incomplete. The missing data
cannot then be reconstructed and is likely to be permanently lost.
If another disk in the array fails when fault tolerance is unavailable during rebuild, a fatal system
error can occur. If this happens, all data on the array is lost. However, in the following cases,
failure of another disk does not lead to a fatal system error:
Failure after activation of a spare disk.
Failure of a disk that is not mirrored to another failed disk (in a RAID 1+0 configuration).
Failure of a second disk in a RAID ADG configuration.
Time required for a rebuild
The time required for a rebuild varies considerably, depending on the following factors:
Priority that the rebuild is given over normal I/O operations
Amount of I/O activity during the rebuild operation.
Rotational speed of the hard disks.
106 Physical disk installation and replacement
Hide quick links:
Advertisement
Table of Contents
