IBM SG24-5131-00 manual 4.2 7135 Disk Failure, Mirrored 7133 Disk Failure

•Reconnect hdisk0, close the casing, and turn the key to normal mode.

•Power on NodeF then verify that the rootvg logical volumes are no longer stale (lsvg -l rootvg).

6.2.4.2 7135 Disk Failure

Perform the following steps in the event of a disk failure:

•Check, by way of the verification commands, that all the Nodes in the cluster are up and running.

•Optional: Prune the error log on NodeF (errclear 0).

•Monitor cluster logfiles on NodeT if HACMP has been customized to monitor 7135 disk failures.

•Mark a shared disk failed through smit ( smit raidiant; RAIDiant Disk

Array Manager -> Change/Show Drive Status -> select the appropriate hdisk -> select the appropriate physical disk -> F4 to select a Drive Status of 83 Fail Drive), or if the disk is hot pluggable, remove the disk.

•The amber light on the front of the 7135 comes on, and can also be seen in SMIT (smit raidiant; RAIDiant Disk Array Manager -> List all SCSI RAID Arrays).

•Verify that all sharedvg file systems and paging spaces are accessible ( df and lsps -a).

•If using RAID5 with Hot Spare, verify that reconstruction has completed to the Hot Spare, then un-mark or plug the failed disk back in. If using RAID1, sync the volume group (syncvg NodeFvg).

•If using RAID5 without Hot Spare, mark the failed disk Optimal (smit raidiant; RAIDiant Disk Array Manager -> Change/Show Drive Status; select the appropriate hdisk -> select the appropriate physical disk -> F4 to select a Drive Status of 84 Replace and Reconstruct Drive).

•Verify that the reconstruction has completed ( smit raidiant;RAIDiant Disk Array Manager -> List all SCSI RAID Arrays).

•Verify that all sharedvg file systems and paging spaces are accessible (df and lsps -a) and that the partitions are not stale (lsvg -l sharedvg). Also verify that the yellow light has turned off on the 7135.

6.2.4.3 Mirrored 7133 Disk Failure

•Check, by way of the verification commands, that all the Nodes in the cluster are up and running.

•Optional: Prune the error log on NodeF (errclear 0).

140 IBM Certification Study Guide AIX HACMP