HyperSwap failover with HyperSwap and LUN switching

In the case of an unplanned outage of a storage server, a HyperSwap® failover is initiated, resulting in near-zero downtime from a user perspective. The figure on the left shows a HyperSwap relationship between all of the disk units in SYSBAS A and SYSBAS B, and in IASP1. The primary access for SYSBAS A and IASP1 is DS8000® M, and the primary access for SYSBAS B is DS8000 N. In addition to being configured for HyperSwap, IASP1 is also configured for LUN level switching to a second IBM® i partition, named IBM i B in the figure.

In order for the HyperSwap failover to be initiated, the following conditions must be met.
  1. All disk units in the ASP group must be configured in a HyperSwap relationship.
  2. All of the HyperSwap relationships in the ASP group must be fully synchronized.
  3. All of the HyperSwap relationships in the ASP group must be going the same direction (some disk units in the ASP group can’t have one DS8000 as the primary while other disk units have the other DS8000 as primary).
HyperSwap Failover with HyperSwap and LUN switching example.

After the HyperSwap failover completes, the primary access for IASP1 and SYSBAS will be the second DS8000, as shown in the figure on the right side. If the HyperSwap failover cannot be completed, the IASP goes into DASD attention.

After the failover occurs and the failed DS8000 comes back online, the HyperSwap relationship must be restarted by entering the following command.

CHGHYSSTS OPTION(*START) NODE(*) ASPDEV(*ALL)

This will restart replication for all ASP devices. The Display HyperSwap Status and Work with HyperSwap Status commands can then be used to monitor the resynchronization process. After resynchronization is completed, a HyperSwap switchover can be initiated if desired.