SAS RAID controller maintenance

Ensure optimal performance of your controller by using these maintenance procedures.

To help avoid controller and disk array problems, use the following tips:

• Perform a normal system shutdown before physically replacing or moving the RAID controller or members of disk arrays. A normal shutdown of the system flushes the controller's write cache and remove dependencies between the controller and the disks. Powering off the PCI slot using concurrent maintenance options in Hardware Service Manager (HSM) has the same effect as it would on a single controller when the Power Down System (PWRDWNSYS) command is used.
  Note: Disks that are failed members of an unprotected (exposed) disk array can be replaced and the disk unit data rebuilt while the system continues to run. No system shutdown is required.
• You can physically move disks from one controller to another. However, if the disks are members of a disk array, be sure to move all the disks in the array as a group. Before attempting a disk movement, ensure that the disk array is not in an unprotected state because of a disk failure. The system or partition must be powered off normally before disks are moved.
• When physically removing disks that are members of a disk array, remove the disks from the auxiliary storage pool (ASP) and then stop RAID on the disk array before removing the disks. This action avoids loss of data and disk-array-related problems the next time that these disks are used. The system or partition must be powered off normally before disks are physically removed.
• Always use the Device Concurrent Maintenance option to remove and replace a disk.
• If the load source disk is part of a disk array and the system fails to IPL because of a suspected disk array problem, IPL the system or partition using D-IPL media (CD/DVD or SAVESYS media). Error Log Analysis, and other tools are available on the Dedicated Service Tools menu to help determine and resolve the problem with the disk array.
• Do not attempt to correct problems by swapping controllers and disks unless you are directed to do so by the service procedures. Use Error Log Analysis to determine what actions to perform, and when appropriate, follow the appropriate Isolation Procedures for problem determination. If multiple errors occur at approximately the same time, look at them as a whole to determine if there is a common cause.
• Do not confuse the cache directory card, which is a small rectangular card with round, button-shaped batteries, for a removable cache card. The nonvolatile write cache memory is integrated into the controller. The write cache memory itself is battery-backed by the large, rechargeable cache battery pack. The cache directory card contains only a secondary copy of the write cache directory and no cache data. Do not remove this card except under specific recovery cases as described in the Isolation Procedures.
• Do not unplug or exchange a cache battery pack without following the procedures as outlined in this section or in the Isolation Procedures. Failure to follow these procedures might result in data loss.