We're trying to catch run away tasks. So the goal is to limit individual job/step CPU time. We have soft caps on our LPARs (still running z/OS 1.7 so we can't use LPAR Groups yet). Our test system has ended up being capped regularly. When that happens we've had problems with GRS communicating through the ring because the test GRS can't get the CPU fast enough. So we've had production jobs time out waiting for a GRS response. Oh, by the way, we only have one production and one test LPAR in the ring, and we are not SYSPlexed. What we've found when the test system hits the cap is that one or two tasks are using an abnormal amount of CPU, and these particular task should probably be running off hours. We want the programmers doing that to "feel our pain" :-). Maybe if their jobs get terminated a couple of times they'll put some efficiencies into their processes, or run outside of prime time. Were also putting CPU limiting factors into our test CICS and test DB2 systems.
I checked out what John McKown had posted about using the IEFUJV exit, and that appears to be what we need. The information in the manual for that exit says that it's the one to use to "validate or assign job TIME and step TIME parameters." That surprised me as I thought the IEFUSI exit would be used for step TIME, but the information in the manual concerning that exit says nothing about using it for the TIME parameter. Tom Kelman Enterprise Capacity Planner Commerce Bank of Kansas City (816) 760-7632 > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] On > Behalf Of Ken Porowski > Sent: Tuesday, July 28, 2009 3:59 PM > To: [email protected] > Subject: Re: Enforcing CPU Time > > Points to consider > > Time limit on the entire job and or step level > Limit based on time of day job is executing vs. time submitted. > Allow TIME= specified at any time or always ignore > Exceptions to the above (they always seem to happen and you don't want > to be on the losing side of a management war). > > What is the real goal, to limit individual jobs or to limit overall CPU > usage by development during prime shift? > A WLM resource group with CPU limits may be appropriate depending on > your exact needs. > > I used to have a UJV exit that required TIME= to be coded for certain > job classes and specific maximums for a given class. TSO submit exit > (IKJEFF10) was used to enforce job classes development could use. > It wasn't perfect, you could bypass IKJEFF10 a number of ways and you > could change your job class after submit through SDSF (Yes I know there > are more complete ways of securing this but it wasn't worth the effort). > Now with only a dozen or so Mainframe developers we can easily say don't > do that and they won't. > > Ken > ***************************************************************************** If you wish to communicate securely with Commerce Bank and its affiliates, you must log into your account under Online Services at http://www.commercebank.com or use the Commerce Bank Secure Email Message Center at https://securemail.commercebank.com NOTICE: This electronic mail message and any attached files are confidential. The information is exclusively for the use of the individual or entity intended as the recipient. If you are not the intended recipient, any use, copying, printing, reviewing, retention, disclosure, distribution or forwarding of the message or any attached file is not authorized and is strictly prohibited. If you have received this electronic mail message in error, please advise the sender by reply electronic mail immediately and permanently delete the original transmission, any attachments and any copies of this message from your computer system. ***************************************************************************** ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
