Automatic Restart Manager planning

z/OS® includes a component known as the Automatic Restart Manager (ARM). ARM offers automated restart capabilities that are attractive in a CICS® TS environment.

ARM is a recovery function that provides improved availability for batch jobs and started tasks by automatically restarting them after they unexpectedly end. In a Parallel Sysplex®, ARM provides additional benefit through its ability to restart registered clients on another MVS™ system image in the Parallel Sysplex if the MVS system image they were originally using fails. Transaction managers, resource managers, and restartable subsystems can be restarted automatically. Any system that require automatic restart must register with ARM. Systems affected by a failure are usually restarted on the same MVS system image, or on a different one if the MVS system image itself has failed.

The cost of computing, productivity, and availability are improved because shared resources and transactions in progress can be recovered and lost function and services restored. The following points apply to the use of ARM:
  • ARM provides job and started task recovery. Transaction or database recovery is the responsibility of the restarted applications.
  • ARM does not start the applications initially (that is, at the first IPL or subsequent IPLs after failures). This is the function of automation or production control products. Interface points are provided through exits, the Event Notification Facility (ENF), and macros.
  • The MVS system image or Parallel Sysplex must have sufficient spare capacity to guarantee a successful restart.
  • Elements that are on an MVS system image that fails are restarted on another MVS system image.

ARM support is part of the overall task of planning for and installing Parallel Sysplex support.

Related reading: For more information about ARM, see z/OS MVS Setting Up a Sysplex.