Lock wait parameters

Use these parameters to modify the time intervals for the server lock wait retry mechanism.

This is provided mainly as a "wake-up" mechanism to ensure that requests waiting for a record lock do not wait indefinitely in certain rare cases where a system failure or structure full condition could cause a lock release notification message to be lost. In addition, this mechanism also ensures that if a CICS task is purged while it has a request waiting for a lock, the waiting request in the server is woken up as soon as the lock wait retry interval expires. The request process then finds that the CICS task is no longer waiting for it, therefore it terminates rather than continuing to wait and hold on to server resources until the lock becomes free.

There are two times involved: a scan time interval and a wait time. The server starts its lock scan interval timing when the first request is made to wait.

This mechanism has very little effect on normal processing and the default lock wait retry parameter values are designed to suit the majority of installations.

LOCKSCANINTERVAL={5|number}
specifies the time interval after which requests waiting for record locks are scanned to check for lock wait timeout.

This affects the overall duration of the lock wait timeout, because a request that starts waiting for a lock during a given scan interval is timed as if from the start of the interval. The lock scan interval should be less than the lock wait interval, and ideally should be such that the lock wait interval is an exact multiple of the lock scan interval.

You can specify this value as a number of seconds or in the time format hh:mm:ss.

The valid range is from 1 (1 second) to 24:00 (24 hours).

This keyword can be abbreviated to LOCKSCANINT.

LOCKWAITINTERVAL={10|number}
specifies the maximum time that a request should wait for a record lock before being reinvoked to retry. The actual time a request waits depends on how far into a scan interval it started its wait. For example, in a server using the scan and wait default intervals, if 4 seconds have already elapsed when a request starts to wait, the maximum time it can wait is 6 seconds. When the server checks the timeout value for the request, it assumes that the request has been waiting for the full scan period. Thus, for the default values, a request is reinvoked to retry after waiting between five and ten seconds.

This value can either be specified as a number of seconds or in time interval format hh:mm:ss.

The valid range is from 1 (1 second) to 24:00 (24 hours).

This keyword can be abbreviated to LOCKWAITINT.



dfha2_lock_wait_parms.html | Timestamp icon Last updated: Sunday, 4 June 2017