Automatic restart management

CICS® uses the automatic restart manager (ARM) component of MVS™ to increase the availability of your systems.

MVS automatic restart management is a sysplex-wide integrated automatic restart mechanism that performs the following tasks:
  • Restarts an MVS subsystem in place if it abends (or if a monitor program notifies ARM of a stall condition)
  • Restarts all the elements of a workload (for example, CICS TORs, AORs, FORs, DB2®) on another MVS image after an MVS failure
  • Restarts CICS data sharing servers in the event of a server failure.
  • Restarts a failed MVS image

CICS reconnects to DBCTL and the z/OS® Communications Server automatically if either of these subsystems restart after a failure. CICS is not dependent on using ARM to reconnect in the event of failure.

The MVS automatic restart manager provides the following benefits:
  • Enables CICS to preserve data integrity automatically in the event of any system failure.
  • Eliminates the need for operator-initiated restarts, or restarts by other automatic packages, thereby:
    • Improving emergency restart times
    • Reducing errors
    • Reducing complexity.
  • Provides cross-system restart capability. It ensures that the workload is restarted on MVS images with spare capacity, by working with the MVS workload manager.
  • Allows all elements within a restart group to be restarted in parallel. Restart levels (using the ARM WAITPRED protocol) ensure the correct starting sequence of dependent or related subsystems.

Restrictions

MVS automatic restart management is available only to those MVS subsystems that register with ARM. CICS regions register with ARM automatically as part of CICS system initialization. If a CICS region fails before it has registered for the first time with ARM, it will not be restarted. After a CICS region has registered, it is restarted by ARM according to a predefined policy for the workload.



dfht2kj.html | Timestamp icon Last updated: Sunday, 4 June 2017