Question:
How to address Unix CP process causing multiple ADR actions to occur for CA Jobtrac Mob Management r11.0?
Answer:
IBM Support has determined that two SMF 15s are being created by the CP process, one with the JFCB indicating a new dataset, the other indicating an existing dataset. Their position is that CA Jobtrac should not trigger on a SMF 15 record for the CLOSE of an existing dataset.
A SCL will need to be setup with a 'D' dependency and '2' in the data set count field to signify not to trigger the job until two SMF type 15 records are created. This will work in conjunction with the Automatic Data Set Recognition table Action: 'DCNT' (decrement count) which will decrement the count in the SCL. When this count is set to zero the job will get released if all other dependencies are met.
Below is an example of SCL that would load the job with a 'D' dependency. Once the decrement value is '0', the 'D' dependency is removed and the job is released.
Example:
TESTJOB1 cannot be submitted until two data sets are created. Two actions are required to schedule the job. First, the ADR table is set to reduce TESTJOB1's data set count by one when each of the data sets is created or updated for TESTJOB1. Second, the Dataset Count (D) field in TESTJOB1's schedule is coded as 2. The following figure shows the previous events scheduled using SCL.
ADR Table:
-------------------------------------------------------------------------- DSN:A.SYS1.TEST.OR.OTHER. TBL.NAME CPU: ACT: DCNT TOD CREATE CREATE TRIGGER TRIGGER THRESHOLD EVENT PROCESS PE EVENT PROCESS PE TESTJOB1 DAILY ************************************************************************** *EVENTNM TIME DATE DEPS RELEASE1 RELEASE2 RELEASE3 RELEASE4 PROCESSP OPD * * * * * * * * * *** TESTJOB1 ASAP * D DAILY 2
As always, please contact CA Technologies support for CA Jobtrac Job Management if you have further questions.