Well we in OMEGAMON have been working on the cpu looping issue lately: Detecting an address space in an infinite loop is not an easy job. Our modern z/OS LPARs often have multiple CPUs and specialty processors like zIIP and zAAP where instructions can be dispatched. In addition Workload Manager will try to distribute processor resources equitably based on the current workload mix and priorities defined in the installation?s policy. Lower priority batch workloads that happen to be looping can easily run under the radar for long periods of time squandering resources.
OMEGAMON XE on z/OS 4.2.0 with the addition of Interim feature 1 has a strategy to help surface these problems. OMEGAMON has had a feature called Bottleneck Analysis for many years. Bottleneck Analysis builds a profile over time through periodic sampling of what execution states are being used by address spaces. These execution states include things like Using CPU, Using zIIP, Using zAAP, Waiting for CPU, Waiting for zIIP, Waiting for zAAP, Using I/O, Waiting for I/O, Waiting for Enqueue, Waiting for HSM, Swapped, etc. A cpu looping address space will reveal itself by populating only the using and waiting states for CPU resources (including zIIP and zAAP). OMEGAMON XE on z/OS 4.2.0 has a new attribute called CPU Loop Index that uses this bottleneck information as its basis. High priority workloads can be reliably detected fairly quickly. The real trick is discriminating between well behaved low priority work that is just starved for attention from low priority work that is looping. OMEGAMON XE on z/OS 4.2.0 Interim Feature 1 provides this discrimination by dynamically extending the observation period required before indicating a likely loop when the ratio of waiting for CPU to Using CPU is high. For more information on OMEGAMON XE on z/OS approach to CPU Loop detection please see the article titled ?Detecting CPU looping address spaces using IBM Tivoli OMEGAMON XE on z/OS version 4.2.0? in the August issue of the z System Advisor at http://www-01.ibm.com/software/tivoli/systemz-advisor/2009-08/. Joe Winterton Release Mgr - OMEGAMON - Development Team 919-224-1328 Cell -914-954-0483 - jose...@us.ibm.com From: Hal Merritt <hmerr...@jackhenry.com> To: IBM-MAIN@bama.ua.edu Date: 12/07/2009 09:48 AM Subject: Re: Detect the loop for batch job Sent by: IBM Mainframe Discussion List <IBM-MAIN@bama.ua.edu> All programs 'loop'. It is what they do. About the only automated solution I can think of would be to set a CPU time limit. -----Original Message----- From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On Behalf Of Joel C. Ewing Sent: Saturday, December 05, 2009 11:20 AM To: IBM-MAIN@bama.ua.edu Subject: Re: Detect the loop for batch job On 12/04/2009 10:14 AM, bjbxd wrote: > Hello List, > We are looking for a tool to detect the loop for batch application, > any suggestion are appreciated. > > My shop is runing z/OS, application is C/C++. > Bob. NOTICE: This electronic mail message and any files transmitted with it are intended exclusively for the individual or entity to which it is addressed. The message, together with any attachment, may contain confidential and/or privileged information. Any unauthorized review, use, printing, saving, copying, disclosure or distribution is strictly prohibited. If you have received this message in error, please immediately advise the sender by reply email and delete all copies. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
