Maximizing access to journal data sets

Where and how you allocate journal data sets affects XRC performance, except in migration mode. XRC journals all copied data, and the capacity of the journal data sets must be enough to support the primary system I/O rate. You do not need journal and control data sets when the XRC session is running in SESSIONTYPE(MIGRATE) mode.

XRC can occasionally fill all of the journal data sets, in which case it must wait for secondary update processing to release journal space before continuing. If the full-journal condition is not relieved within a reasonable amount of time, XRC suspends the session, and then restarts it. If this happens, you may need to enlarge the journal data sets. Follow the procedure that is documented in Specifying XRC journal, control, and state data sets to enlarge the journal data sets.

Perform the following steps to enlarge the journal data sets.

  1. Issue the XSUSPEND command with the TIMEOUT option to suspend the session if it is still active.
  2. If the session is coupled, issue the XCOUPLE command with the PURGE option to uncouple it.
  3. Restart the XRC session with the XSTART command with SESSIONTYPE(MIGRATE). (See XSTART–Starting a session).
  4. Again, issue the XSUSPEND command with the TIMEOUT option to suspend the session.
  5. Enlarge the size of the journal data sets or allocate additional ones.
  6. Restart the XRC session using the XSTART command with SESSIONTYPE(XRC).
  7. Perform resynchronization using the normal procedures for your installation.
Note: The state and control data sets can reside on the same volume. The control data set, however, should not reside on the same volumes as the journal data sets.

For additional information about how to determine the correct journal data set size, and about optimizing performance, refer to Specifying the journal data sets.

Parent topic: Optimizing XRC performance