CICS workload routing and management

CICS® provides a number of capabilities for automatically and efficiently distributing transactions within a Parallel Sysplex®.

When work arrives in CICS a number of things can happen:
  • The request can run entirely within a single CICS region.
  • But, typically, in a high-availability Parallel Sysplex, some or all of the request is routed.
If you remove all affinities from your CICS applications, any transaction is eligible to run in any available CICS region. However, to maintain availability, transactions must be automatically routed as quickly as possible to the most appropriate CICS region.
There are three main components that are involved in managing a CICS workload in a Parallel Sysplex:
  • CICSPlex® SM

    You provide CICSPlex SM with the CICS-specific workload specifications and definitions that it needs to support both workload management and workload separation in the CICSplex. These workload specifications and definitions ensure that CICSPlex SM has all the necessary information about the transactions that comprise the CICSplex workload, and about the application-owning regions that are available to process the work.

  • A CICS dynamic routing program

    You need a dynamic transaction routing program that can use the services of CICSPlex SM and route transactions to the appropriate application-owning regions. If you implement CICSPlex SM, you can use the dynamic transaction routing program that it provides.

  • The z/OS® workload manager

    You provide the workload manager with a service definition, which z/OS needs to ensure that it makes the necessary resources available to the CICS regions and associated subsystems that ultimately process the user transactions. The z/OS service definition must cover the workload for the entire sysplex, so that z/OS can allocate resources for all types of work; online, batch, and system-related.

    The service definition is primarily needed for goal-oriented workload management.

For more information about managing and monitoring your CICS workload, see Managing and monitoring your CICS workload.