Question & Answer
Why has the CPU consumption increased by over 100% in the past 10 days in my CICS Transaction Server for z/OS (CICS TS) region? This region is now being cycled on a weekly basis when previously it was cycled daily. I have a dump taken when CPU usage was 100%.
I am using INNOVATION ACCESS METHOD (IAM) from INNOVATION DATA PROCESSING.
Looking at the system dump, enter command SYSTRACE PERFORDATA from IPCS option 6).
In the CPU breakdown by ASID, the TCB Adr 009CC4D8 stood out as taking a lot of CPU time:
ASID Jobname TCB Adr # of DSPs Time 0184 <cicsjob> 009C9E88 6 0.000170 0184 <cicsjob> 009BD488 24 0.000874 0184 <cicsjob> 009CC4D8 772 0.154085 <<<<< ------------ 0.155131
and the 'Interrupt PSW' D2BDCCB0 (or 52BDCCB0) for TCB Adr 009CC4D8 shows a Count of 574 times!
SRB/TCB IASN Interrupt PSW Count -------- ---- -------- -------- -------- 009CC4D8 0184 078D2000 D2BDCCB0 574
Next, enter option 1 from the IPCS primary option menu to Browse the dump data set starting at the 'Interrupt PSW' address then scroll backwards until you find the eyecatcher of the failing module. In this case, the 'Interrupt PSW' D2BDCCB0 (or 52BDCCB0) is pointing to module IAMACCXM at offset 2F18. The eyecatcher for IAM module IAMACCXM contains a date of 6/1/73 ver 9.1/01P .¢09.0:
52BD9D98 e .02..03m e 52BD9DA0 e .03..03*.03....q e 52BD9DB0 e ...q..w?..x....d e 52BD9DC0 e .IAMACCXMPROPRIE e 52BD9DD0 e TARY CODE INNOVA e 52BD9DE0 e TION DATA PROCES e 52BD9DF0 e SING 6/1/73 -- V e 52BD9E00 e ER.9.1/01P .!09. e
The IAM fix for this problem is Zap ID p-91.0105. Contact INNOVATION DATA PROCESSING for IAM Support if you need additional assistance.
CICS/TS CICSTS CICS TS CICS Transaction Server
18 September 2018