|
Intel® RAID Controller SRCU32, Intel® RAID Controller SRCU42L, Intel® RAID Controller SRCS14L, & Intel® RAID Controller SRCZCR with firmware versions 2.42.00-R072 or 2.42.01-R078 that are used in a RAID 10 configuration may experience a rebuild checkpoint failure, if a drive fails and is detected again at the next reboot.
Intel introduced “check-pointing” during build and rebuild in firmware versions 2.42.00-R072 and 2.42.01-R078 (global release 1.10.8 & 1.10.9). Rather than restarting a rebuild during system reboot, check-pointing tracks rebuild progress and allows a rebuild to restart where it stopped should a reboot occur prior to the completion of the rebuild. Check pointing
is used in RAID 1, RAID 4, RAID 5, and RAID 10 configurations.
However, if one drive of a RAID 10 fails during operation, a rebuild has not started, and the failed drive is left in the system during a reboot, the failed drive may be detected and become part of the array again. In this situation a rebuild will start and immediately stop even though the missing mirror component is no longer in synch with the other array members.
This issue is not observed with RAID 1, RAID 4, or RAID 5 configurations.
 Technical Advisory TA-743-1 [PDF]
File Name: ta7431.pdf
Size: 28,019 bytes
Date: January 2005
File Revision: TA-743-1
This applies to:
| 
