"McKown, John" <[EMAIL PROTECTED]> wrote in message news:<[EMAIL PROTECTED]>... > > -----Original Message----- > > From: IBM Mainframe Discussion List > > [mailto:[EMAIL PROTECTED] On Behalf Of Kelman, Tom > > Sent: Friday, October 12, 2007 12:04 PM > > To: [email protected] > > Subject: Are there tasks that don't play by WLM's rules > > > > I am cross posting this to the IBM-Main and the MXG listservs. > > > > We are about to set the soft cap on our systems. I have been informed > > by my system programmers that in the past they have seen tasks that > > don't play by WLM's rules when it comes to resource > > utilization. One is > > the Direct-Connect (NDM) task. In the past, before I came to this > > location, they saw Direct-Connect take CPU resources and affect CICS > > processing when the system was heavily used even though Direct-Connect > > had an importance level lower than the CICS regions. We are > > running NDM > > at an importance level of 4 with a velocity goal of 20. Our CICS > > regions are at an importance level of 1 with a velocity goal > > of 50. The > > CICS transactions are run at an importance level of 1 with a response > > time goal of 80% within 0.5 seconds. > > > > Has anyone else experienced this with Direct-Connect or any other > > application? That is have they experienced the application > > taking over > > the system even though it has a low importance level? This > > is important > > to us as we want our CICS processing to be one of the last > > things to be > > hurt if we hit the soft cap. > > > > Tom Kelman > > > > Commerce Bank of Kansas City > > > > (816) 760-7632 > > In a sense, yes. I've had test batch perform better than production > batch, despite a lower velocity and importance. The reason was that > there was a single job running in the test service class and many jobs > running in production. In this case, the PI of the test service class > was greater than the PI of the production one, so WLM gave CPU cycles to > the single test job. > > I don't have this particular problem with production CICS because I have > marked the production CICS service class as CPU CRITICAL. This means > that WLM will NEVER steal CPU cycles from CICS to help other, lower > performance, work even if the CICS PI is smaller than the other work's > PI. >
This is not exactly correct: CPU critical will prevent the task's Dispatching Priority from dropping below the DP of any lower IMP task. WLM will steal cycles from an overperforming CICS to help lower IMP tasks. Kees. ********************************************************************** For information, services and offers, please visit our web site: http://www.klm.com. This e-mail and any attachment may contain confidential and privileged material intended for the addressee only. If you are not the addressee, you are notified that no part of the e-mail or any attachment may be disclosed, copied or distributed, and that any other action related to this e-mail or attachment is strictly prohibited, and may be unlawful. If you have received this e-mail by error, please notify the sender immediately by return e-mail, and delete this message. Koninklijke Luchtvaart Maatschappij NV (KLM), its subsidiaries and/or its employees shall not be liable for the incorrect or incomplete transmission of this e-mail or any attachments, nor responsible for any delay in receipt. Koninklijke Luchtvaart Maatschappij N.V. (also known as KLM Royal Dutch Airlines) is registered in Amstelveen, The Netherlands, with registered number 33014286 ********************************************************************** ---------------------------------------------------------------------- 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
