We have a TX Series v.7 transaction running on a AIX with a loop in a COBOL program; we are analyzing this loop.
While we are fixing the issue we need to remove the process manually. My experience is 100% CICS;
CICS use the ICVR system initialization parameter ( SIT ) value as the runaway time limit for transaction.
I think that TX Series must have a similar parameter to limit the use of cpu or a time limit.
Somebody knows about a TX Series v7 parameter used to stop the execution of cpu high consuming transaction ?
Thanks in advance
CICS - ICVR Definition
- specifies the default runaway task time interval in milliseconds as a decimal number. You can specify zero, or a number in the range 500 through 2 700 000, in multiples of 500. CICS rounds down values that are not multiples of 500. This is the RUNAWAY interval used by transactions defined with RUNAWAY=SYSTEM (see ../../dfha4/resources/transaction/dfha4_attributes.html for further information). CICS may purge a task if it has not given up control after the RUNAWAY interval for the transaction (or ICVR if the transaction definition specified RUNAWAY=SYSTEM). If you code ICVR=0, runaway task control is inoperative for transactions specifying RUNAWAY=SYSTEM in their transaction definition (that is, tasks do not get purged if they appear to be looping). The ICVR value is independent of the ICV value, and can be less than the ICV value. Note that CICS runaway task detection is based upon task time, that is, the interval is decremented only when the task has control of the processor. For information about commands that reinitialize the ICVR value, see Investigating loops that cause transactions to abend with abend code AICA.