Previous topic |
Next topic |
Contents |
Contact z/OS |
Library |
PDF
SQA/CSA thresholds z/OS MVS Initialization and Tuning Guide SA23-1379-02 |
||||||||||||||||||||||||||
Ensuring the appropriate size of the extended SQA and extended CSA storage is critical to the long-term operation of the system. If the size allocated for SQA is too large, the thresholds for the IRA100E and IRA101E messages will not be met and CSA can become exhausted and cause a system outage. One way to avoid this problem is to allocate the minimum SQA required or allow for some CSA conversion so that the storage thresholds that trigger the IRA100E and IRA101E messages are based only on the remaining CSA. If the size allocated for extended SQA is too small or is used up very quickly, the system attempts to use extended CSA. When both extended SQA and extended CSA are used up, the system allocates space from SQA and CSA below 16 megabytes. The allocation of this storage could eventually lead to a system failure.
The following conditions may be responsible for a shortage of SQA/CSA:
Note: IBM® recommends that
you do not change the storage thresholds set by the system. Setting
the thresholds too low can hamper the ability of the system to recover
from storage shortages and may result in unscheduled system outages.
Setting the thresholds too high can cause IRA100E, IRA101E, and IRA102I
messages to be issued excessively. If you do change the storage thresholds,
you should be sure that the threshold is not being reached because
of a problem in the system or inadequate allocation of CSA/SQA.
The
SQA/CSA threshold levels are contained in the IGVDCLIM CSECT in load
module IEAIPL04. Table 1 shows the
offsets of the threshold values.
If you change the default thresholds, make sure that the new sufficient threshold values are at least x'1000' larger than the insufficient values. The new values become effective at the next IPL. |
Copyright IBM Corporation 1990, 2014
|