Jim, many thanks for the comprehensive reply. Very helpful. And thanks to all the others who replied. Werner
IMD-Gesellschaft für Informatik und Datenverarbeitung mbH Augustaanlage 66 68165 Mannheim Germany Tel: +49.621.457-4885, Fax: -4046 E-Mail: [EMAIL PROTECTED] IBM Mainframe Discussion List <IBM-MAIN@BAMA.UA.EDU> schrieb am 10.01.2008 16:36:00: > NOTICE: > All information in and attached to the e-mail(s) below may be > proprietary, confidential, privileged and otherwise protected from > improper or erroneous disclosure. If you are not the sender's > intended recipient, you are not authorized to intercept, read, > print, retain, copy, forward, or disseminate this message. If you > have erroneously received this communication, please notify the > sender immediately by phone (704-758-1000) or by e-mail and destroy > all copies of this message (electronic, paper, or otherwise). Thank you. > > Werner, > > In a word, yes. For more words, read on. The SRM constant for TYPE72 > processing is determined by the number of online general purpose engines > an LPAR has when it is IPLed, not the number of possible engines > available to it. So, your LPAR IPLed with three engines on a 2094-720 > thinks it is on a 2094-703. > > Additionally, if you vary engines online or offline (either manually or > using IRD), the SRM constant does not change. This means that an LPAR > can appear to be more than 100% busy under the right (wrong) > circumstances. Take, for example, your LPAR that was IPLed with 3 > engines for an SRM constant of 27539 (using your numbers; I haven't > looked them up - I trust you). If seventeen engines are varied online, > that means that the systems thinks it has 20 engines which can each > deliver 27539 services units per second, instead of the 20126 service > units per second (again, using your numbers) it would believe it had if > it had been IPLed with all twenty engines online. In table form: > Machine SU/SEC Total SU Total SU Total SU > > at IPL w/3 engines w/ 20 > engines > 2094-703 27539 82617 82617 550780 > 2094-720 20126 402520 61560 405520 > > As you can see, the reported service units don't match up well at all > with the actual service units when the number of engines being used > doesn't match the number of engines the LPAR was IPLed with, or when you > try to compare an LPAR against an entire machine. > > This is called a "reporting opportunity". I would not advise using an > individual LPAR's view of service units for reporting anything that > happens in the entire machine; I would only use it for reporting things > in that LPAR. I do not care for the service unit view that the RMF > records give us - but I do have to deal with it. > > /shameless plug on > I will be presenting the same paper at Share in Orlando that I presented > at MXG2007, "The Myth of MSU," at 8AM Thursday morning. In this paper I > go into some detail about what I found investigating the TYPE70 and > TYPE72 records and what they actually tell us when specialty engines > (zIIPs and zAAPs) and IRD (Intelligent Resource Director) are taken into > account. I personally found the results very surprising when I looked > into it after my management started asking questions. That's why I > wrote the paper. > /shameless plug off > > Hope this helps, > Jim Horne > Lowe's Companies, Inc. ---------------------------------------------------------------------- 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
