Shutting down CICS

This topic describes the three types of CICS® system shutdown (normal, immediate, or uncontrolled) and the events that cause them. For information about CICS shutdown, see the CICS Recovery and Restart Guide.

After a normal shutdown, it is possible to warm start CICS. After an immediate or an uncontrolled shutdown, an emergency restart or a cold start must be performed.

CICS XRF systems

When an XRF active CICS region is terminated abnormally, the alternate CICS region normally completes initialization and takes over. However, the alternate CICS region also terminates if the user has initiated a normal shutdown of the active CICS region and has not specified that takeover is to take place.

Normal shutdown (PERFORM SHUTDOWN)

Normal shutdown is initiated by the master terminal operator or by an application program, and is accomplished in phases. (In comparison, immediate shutdown is accomplished by termination processing.)

First stage of normal shutdown

During the first stage of CICS normal shutdown, all terminals are active and all CICS facilities are available. The following actions take place concurrently:

The first shutdown stage is complete when the last of the programs specified in the first part of the shutdown PLT has run and all user tasks are complete.

Second stage of normal shutdown

During the second stage of shutdown, terminals are not active, and no new tasks are allowed to start. The following processing takes place:

  1. User-written programs listed in the second part of the shutdown PLT (if any) are executed sequentially. These programs cannot communicate with terminals, or make any request that would cause a new task to start.
  2. All currently open CICS files are now closed.
  3. The transient data CI buffer and the temporary storage buffers are flushed.
  4. CICS writes the following information to the global catalog:
  5. Transient data is terminated.
  6. A dump is taken, if one is required.
  7. If TAKEOVER was specified on the command to shut down an XRF CICS region, a "signoff abnormal" request is made from the CICS availability manager (CAVM).
  8. The local and global catalogs are closed.
  9. The following message is issued:
    DFHKE1799 applid TERMINATION OF CICS IS COMPLETE
  10. CICS completes some internal processing, then returns control to MVS™.

Immediate shutdown (PERFORM SHUTDOWN IMMEDIATE)

During immediate shutdown of CICS, possibly requested by the master terminal operator or an application program, processing is different from a normal shutdown in the following important ways:

To preserve data integrity, the next initialization of CICS must be an emergency restart. If the next initialization of CICS specifies START=AUTO, there will be an emergency restart.

The processing involved in immediate shutdown is described as CICS system termination processing. (In comparison, normal shutdown involves quiesce processing.)

Unlike processing, controls are not exercised to ensure that resources and services remain available as long as they are needed. One consequence of this is that transaction and CICS system abends can occur during immediate shutdown. Thus, if a task tries to use a resource that has already been terminated, the task abends. Then dynamic transaction backout is invoked, and that might also fail because it could also try to use a resource that has been terminated.

In addition, if CICS system termination processing is delayed significantly, tasks in the system waiting for input from terminals that are no longer available are likely to extend beyond the period for deadlock timeout specified in the DTIMOUT option of the TRANSACTION definition.

First stage of immediate shutdown

During the first stage of an immediate shutdown, the following processes take place:

  1. The system termination task drives the collection of termination statistics.
  2. If there is a terminal associated with the event that caused the immediate shutdown, a message is sent to inform the operator that CICS is terminating.
  3. If the shutdown request has arrived by transaction routing, the associated terminal is freed.
  4. Terminal input is no longer accepted.
  5. The Front End Programming Interface (FEPI) is requested to shut down immediately. Unless SDTRAN=NO or NOSDTRAN was specified, the shutdown task starts the specified shutdown transaction (the default is CESD). CESD manages the purging of long-running user tasks.

Second stage of immediate shutdown

During the second stage of an immediate shutdown, the following processing takes place:

  1. Transient data is terminated.
  2. A dump is taken, if requested.
  3. Interregion sessions are terminated.
  4. If CICS is signed on to the CICS availability manager (CAVM), a "signoff abnormal" request is made from CAVM.
  5. The local catalog and global catalog are left to be closed by the operating system.
  6. The following message is issued:
    DFHKE1799 applid TERMINATION OF CICS IS COMPLETE
  7. CICS completes some internal processing, then returns control to MVS.

Uncontrolled shutdown

An uncontrolled shutdown of CICS can be caused by a power failure, a machine check, or an operating system failure.

In each case, CICS cannot perform any shutdown processing. In particular, CICS does not write a warm keypoint or a warm-start-possible indicator to the global catalog.

To preserve data integrity, the next initialization of CICS must be an emergency restart. If the next initialization of CICS specifies START=AUTO, there will be an emergency restart.

Related concepts
Starting up CICS
Controlling CICS operation
[[ Contents Previous Page | Next Page Index ]]