Increasing the CICS region size

If all other factors in a CICS® system are kept constant, increasing the region size available to CICS allows an increase in the dynamic storage areas.

Changes to MVS™ and other subsystems over time generally reduce the amount of storage required below the 16MB line. Thus the CICS region size may be able to be increased when a new release of MVS or non-CICS subsystem is installed.

To get any further increase, operating-system functions and storage areas (such as the local shared queue area, LSQA), or other programs must be reduced. The LSQA is used by VTAM® and other programs, and any increase in the CICS region size decreases the area available for the LSQA, SWA, and subpools 229 and 230. A shortage in these subpools can cause S80A, S40D, and S822 abends.

If you specify a larger region, the value of the relevant dsasize system initialization parameter must be increased or the extra space is not used.

How implemented

The region size is defined in the startup job stream for CICS. Other definitions are made to the operating system or through operating-system console commands.

To determine the maximum region size, determine the size of your private area from RMF™ II or one of the storage monitors available.

To determine the maximum region size you should allocate, use the following formula:

The remaining storage is available for the CICS region; for safety, use 80% or 90% of this number. If the system is static or does not change much, use 90% of this number for the REGION= parameter; if the system is dynamic, or changes frequently, 80% would be more desirable.

Note:
You must maintain a minimum of 200KB of free storage between the top of the region and the bottom of the ESA high private area (the LSQA, the SWA, and subpools 229 and 230).

How monitored

Use RMF, the RMFMON command on TSO for additional information. For more information about RMF see Resource measurement facility (RMF) or the MVS RMF User’s Guide.

Related tasks
MVS and DASD: improving performance
Reducing MVS common system area requirements
Splitting online systems to improve availability
Making CICS nonswappable
Using job initiators
Tuning the region exit interval (ICV)
Using LLA (MVS library lookaside)
DASD tuning

Related reference
Storage requirements for a CICS region
[[ Contents Previous Page | Next Page Index ]]