Cluster repository disks
PowerHA® SystemMirror® uses a shared disk as a central repository for managing the configuration of the cluster. This disk must be accessible by all of the nodes in the cluster by using the standard or stretched cluster deployment method, or must be accessible by all nodes in a site by using the linked cluster deployment method.
Use the System Management Interface Tool (SMIT) interface to manage and configure the cluster repository disk and the backup repository disks.
You must have at least one active repository disk per cluster for standard clusters and stretched clusters. You can identify up to six backup repository disks per cluster for standard clusters and stretched clusters. You must have one active repository disk per site for linked clusters. You can identify up to six backup repository disks per site for linked clusters.
With Cluster Aware AIX® the cluster repository disk is used for the following purposes:
- Cluster-wide configuration management
- Cluster messaging and heartbeating. The repository disk is used as another redundant path of communication between the nodes.
You must have at least 512 MB and no more than 460 GB of disk space that is allocated for the cluster repository disk.
Verify that the disk you select as the repository disk does not have a reservation policy. To
check the disks reservation policy, run the lsattr -El hdisk - a reserve_policy
command. To change the disk reservation policy to no_reserve, run the
chdev –a reserve_policy=no_reserve –l hdisk command.
After you assign a disk as the repository disk, the disk cannot be used for any other purposes. Verify that the disk you use as the repository disk does not contain any user data. When the disk is identified as a repository disk by PowerHA SystemMirror, all information on the disk is erased.
When you plan to use the disks as repository disks, you must plan for backup or replacement disks, which can be used in case the primary repository disk fails. The backup disk must be the same size and type as the primary disk, but might be in a different physical storage disk. Update your administrative procedures and documentation with the backup disk information. The cluster tolerates the loss or failure of the repository disk. Critical operations continue without a repository disk. However, you must quickly correct the problem with the repository disk. If you are using PowerHA SystemMirror 7.2, or earlier, you must manually replace the failed repository disk with a repository disk from the backup list. You can also replace a working repository disk with a new disk to increase the size or to change to a different storage subsystem.
You can configure PowerHA SystemMirror 7.2, or later, to use the Automatic Repository Disk Replacement (ARR) function in Cluster Aware AIX (CAA) during the failure of an active repository disk. The ARR function automatically replaces the failed repository disk with a disk from the backup repository disks. The first backup repository disk in the list replaces the failed repository disk.
You should configure all repository disks with AIX native Mulitple Path I/O (MPIO). The AIX MPIO enables PowerHA SystemMirror to monitor and respond better to disk failures by using repository disks.