z/OS DFSMShsm Implementation and Customization Guide
Previous topic | Next topic | Contents | Contact z/OS | Library | PDF


How secondary host promotion works

z/OS DFSMShsm Implementation and Customization Guide
SC23-6869-01

When a primary or SSM host becomes disabled, all DFSMShsm hosts in the HSMplex are notified through XCF. Any host that is eligible to perform the functions of the failed host will attempt to take over for the failed host. The first host that successfully takes over for the failed host becomes the promoted host. There is no means available for assigning an order to which hosts take over the functions of a failed host.
Note: Secondary host promotion is designed to occur when the primary host fails or becomes unexpectedly disabled. To cause secondary host promotion during a normal shutdown of DFSMShsm, issue the STOP command with the PROMOTE or DUMP parameters.

If an original host is both a primary and an SSM host, its responsibilities can be taken over by two separate hosts.

Example: If a secondary host specifies

SETSYS PROMOTE(PRIMARYHOST(YES) SSM(NO))

and a different secondary host specifies

SETSYS PROMOTE(PRIMARYHOST(NO) SSM(YES)),

then it is possible for each host to take over part of the original failed host’s work.

Likewise, if a secondary host is eligible to be promoted for both primary and SSM host responsibilities, then it can be promoted for two separate hosts.

Example: Host A, the primary host, and host B, an SSM host.

If the promoted host itself fails, then any remaining host that is eligible for promotion will take over. If additional failures occur, promotion continues until there are no remaining hosts that are eligible for promotion.

If a secondary host fails while it is promoted for an original host and there are no remaining active hosts eligible for promotion, then of any of the secondary hosts that become reenabled before the original host does, only that host that was last promoted for the original host can become the promoted host.

Rule: For secondary host promotion to work at its highest potential, do not use system affinity. All systems must have connectivity to all storage groups. If system affinity is used, then storage groups that are only associated with one system would not be processed when that system was not available.

Go to the previous page Go to the next page




Copyright IBM Corporation 1990, 2014