Selecting data sets

Data sets are selected during the Activation phase by performing a catalog look-up using the Source DSN values provided in the Define zDMF Group panel (refer to Creating a group definition). Uncataloged data sets will fail activation.

The Owning zDMF system (the host where the Group Activate command was submitted) performs the catalog look-up. When choosing the Owning system, select the system that possesses the most paths and that performs most of the I/O operations to the data that you are moving. Data sets catalogued in multiple catalogs will be migrated only from the perspective of the Owning system's master catalog.

Note: When migrating multiple zDMF groups for a common pool of source volumes, it is strongly recommended that all these groups be activated on the same LPAR (have the same zDMF Owning system). Otherwise, it will result in two owning systems sharing a migration source volume, which will cause Activation to fail.

Important: Source Data Set Catalogued in Multiple Catalogs: All master catalogs (MCATs) should have equivalent alias definitions, with each alias entry pointing to the same user catalog (usercat) across all MCATs within the SYSPLEX or shared storage complex. The risk to a migration is that zDMF identifies the data sets to be migrated through a catalog search performed on the zDMF Owning system, the system where the zDMF Monitor is used to perform the Group ACTIVATE. If the MCATs do not have synchronized aliases, then there is a risk of having more than one catalog entry for a given data set. The zDMF ACTIVATE phase recognizes only the data set associated with the catalog known to the Owning system. MCAT alias synchronization errors should be resolved as part of your migration planning and preparation.