Administering virtual disks

Hot spares

``Hot spares'' are pieces which are allocated to a virtual disk and are used if one of the other pieces starts to fail. If I/O to a piece starts to fail, the hot spare piece is brought online, and data from the failing piece is moved to it (or restored from redundancy check data stored on the other disks in the array). When the data transfer is complete the hot spare takes the place of the failed piece. This ensures that the system is kept operational even if a disk fails under use.

If a drive in a RAID 1 (mirror) or RAID 5 virtual disk fails, the data availability is reduced immediately. The virtual disk is not protected from additional drive failures until the failed component is repaired. If a disk failure occurs in a virtual disk with a hot spare, data is automatically recreated on the hot spare. When the recreation is complete, the virtual disk operates at its original level of data availability.

If a hot spare is not configured and a disk fails in an array, you can reconfigure the failed piece so it resides on a functional disk.

In the event of a disk failure, there may be some I/O performance degradation while data recovery takes place. Performance returns to its original level after the hot spare has been brought online.

It might be useful to use a single disk drive for all your hot spare pieces.

NOTE: The Virtual Disk software does not support hot swapping of disks. Hot swapping allows a physical disk to be removed from an array, and another disk to be inserted to replace it.

See also:

Next topic: Clusters
Previous topic: Disk arrays and data striping

© 2003 Caldera International, Inc. All rights reserved.
SCO OpenServer Release 5.0.7 -- 11 February 2003