Writing a distributed routing program

You can use a distributed routing program to route different types of request in CICS®, including inbound web services and non-terminal START requests.

The distributed routing program is named on the DSRTPGM system initialization parameter in the routing and target CICS regions. You can use a distributed routing program to route these requests:
  • CICS business transaction services (BTS) processes and activities
  • Non-terminal-related EXEC CICS START requests.

    For information about which non-terminal-related START requests are eligible for distributed routing, see Routing transactions invoked by START commands.

  • Inbound web service requests
You cannot use the distributed routing program to route these requests:
  • Transactions initiated from user terminals
  • Transactions initiated by terminal-related EXEC CICS START commands
  • Program-link requests

To route these requests, you must use the dynamic routing program named on the DTRPGM system initialization parameter. How to write a dynamic routing program is described in Writing a dynamic routing program. The dynamic routing program and the distributed routing program can be the same program.

If you use CICSPlex System Manager (CICSPlex SM) to manage your CICSplex, you can use the routing program, EYU9XLOP. This program supports workload balancing and workload separation. You can define which regions in the CICSplex can participate in the workload, and define any transaction affinities that govern the regions to which particular requests must be routed. For more information about workload management in CICSPlex SM, see Configuring workload management.