Re: increased CPU usage z/os v1r13
On 8/01/2013 22:04, Shane Ginnane wrote: I might refer you to a thread on the RMF spreadsheet reporter from a few weeks back. Andrew ... ? Thanks Shane :-) If you need a tool to read the SMF data, you could download the 30 day trial of EasySMF from: http://www.smfreports.com/ My suggestion would be to set the time interval to cover a few days before and after the change, then look at the charts for: - System CPU Busy - RMF Service by Service Class - RMF Service by Report Class and see whether there are any obvious changes. Depending what you see you can use the Job Status reports to see more detail about the jobs in that service/report class. The Job Profile and Job History reports can show you information about specific address spaces and jobs over time. However, you say you are not sure what the CPU peaks were before migrating. If it was already peaking at or close to 100% the problem might be more related to the WLM policy. At 100% busy, subtle WLM changes can make a big difference between everybody being happy and everybody complaining. If you regularly have multiple service classes missing goals, any changes to how WLM choses donors and receivers in 1.13 might trigger problems. (I don't know whether there were such changes, but I wouldn't be surprised.) For WLM, you could start with the Performance Index report. Another possibility is hiperdispatch, if that has been newly enabled with 1.13. I suspect hiperdispatch would have similar effects to moving to fewer, faster processors which has been well documented. Fewer, faster processors makes the WLM policy more important and problems tend to show up as complaints about response times. You should definitely look at whether the CPU usage has increased and how much, and whether there are specific problems that caused it. However, keep an open mind to other factors like WLM. Regards Andrew Rowley -- Andrew Rowley Black Hill Software Pty Ltd and...@blackhillsoftware.com +61 413 302 386 -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: increased CPU usage z/os v1r13
On Mon, 7 Jan 2013 14:09:32 -0500, Crabtree, Anne D wrote: >Since bringing up z/os v1r13 in late October, our CPU stays at 100%, batch >jobs are backing up and people are complaining. (Migrated from z/os v1r11) > >...I can't say for sure what kind of percentage the CPU was peaking at for >1.11, but I know that we did not get these phone calls complaining about >response time! You say in a later post that you bill for CPU usage - so we can only assume you have all the (historical) data you need. Presumably you don't have a SAS/MXG pdb, else you wouldn't be asking. You have to determine where the extra (apparent) CPU burn is - by workload, or service class, whatever. Find the variation, then you know where to concentrate - might even be uncaptured, in which case it really might be worth looking at things like exits. I'm currently working at a somewhat challenged environment in the middle of a 1.11 to 1.13 rollout - we see some issues, but nothing directly attributable to 1.13 eating more machine in peak (daylight) hours as you later describe. I might refer you to a thread on the RMF spreadsheet reporter from a few weeks back. Andrew ... ? Shane ... -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: increased CPU usage z/os v1r13
Hi I don't remember to see anything like this, but as we changed accidentally some LE runtime options, it had a very bad impact on the CPU usage of the XPLINK applications. Here maybe the "D CEE,ALL" would help. First I would try to find out who are the main contributors for this 100% load. If nothing better, the RMF III workflow exception report can point out the critical applications On 07.01.2013 20:19, Crabtree, Anne D wrote: Sorry, kind of misleading in original post because the CPU only stays at 100% during the day, mostly 9am(ish) - 3pm(ish) with an obvious drop-off noon-1pm. I definitely worked on IEFACTRT but it was because of rate changes (we bill for CPU usage) so I know it's been reassembled! You've made me paranoid that I've already asked this question and if I did, blame it on sometimer's disease..lol Anne D. Crabtree System Programmer WV Office of Technology Data Center 1900 Kanawha Blvd East Bldg 6, Room B-110 Charleston, WV 25305 (304)558-5914 ext 58292 (304)558-1441 fax -Original Message- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf Of Brian France Sent: Monday, January 07, 2013 2:16 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: increased CPU usage z/os v1r13 Migrated back in October, z/OS 1.11 to 1.13. Saw nothing of the like. For some reason I kinda remember another email on this subject sometime in the last 6 months or so. Please don't hold me this to this but an exit may have needed reassembled and I wanna say it was iefactrt that was not working right. You might be able to search the archives for it. Sorry best this feeble brain can do. On 1/7/2013 2:09 PM, Crabtree, Anne D wrote: Since bringing up z/os v1r13 in late October, our CPU stays at 100%, batch jobs are backing up and people are complaining. (Migrated from z/os v1r11) I've looked for information on increased CPU for 1.13 but haven't found much. I see now that there is a product you can download to predict increased CPU when migrating, but, unfortunately, I didn't realize that until now. I can't say for sure what kind of percentage the CPU was peaking at for 1.11, but I know that we did not get these phone calls complaining about response time! I just wondered if anyone else experienced a significant increase in CPU usage going to 1.13? Anne D. Crabtree System Programmer WV Office of Technology Data Center 1900 Kanawha Blvd East Bldg 6, Room B-110 Charleston, WV 25305 (304)558-5914 ext 58292 (304)558-1441 fax -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- Brian W. France Systems Administrator (Mainframe) Pennsylvania State University Administrative Information Services - Infrastructure/SYSARC Rm 25 Shields Bldg., University Park, Pa. 16802 814-863-4739 b...@psu.edu "To make an apple pie from scratch, you must first invent the universe." Carl Sagan -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: increased CPU usage z/os v1r13
You might search the archives for CEE parm RTEREUS and see if that might apply to your environment. One post mentioned that "Without RTEREUS the wall clock time was doubled or tripled." HTH, Greg Shirey Ben E. Keith Company -Original Message- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf Of Crabtree, Anne D Sent: Monday, January 07, 2013 1:19 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: increased CPU usage z/os v1r13 Sorry, kind of misleading in original post because the CPU only stays at 100% during the day, mostly 9am(ish) - 3pm(ish) with an obvious drop-off noon-1pm. I definitely worked on IEFACTRT but it was because of rate changes (we bill for CPU usage) so I know it's been reassembled! You've made me paranoid that I've already asked this question and if I did, blame it on sometimer's disease..lol Anne D. Crabtree -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: increased CPU usage z/os v1r13
Sorry, kind of misleading in original post because the CPU only stays at 100% during the day, mostly 9am(ish) - 3pm(ish) with an obvious drop-off noon-1pm. I definitely worked on IEFACTRT but it was because of rate changes (we bill for CPU usage) so I know it's been reassembled! You've made me paranoid that I've already asked this question and if I did, blame it on sometimer's disease..lol Anne D. Crabtree System Programmer WV Office of Technology Data Center 1900 Kanawha Blvd East Bldg 6, Room B-110 Charleston, WV 25305 (304)558-5914 ext 58292 (304)558-1441 fax -Original Message- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf Of Brian France Sent: Monday, January 07, 2013 2:16 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: increased CPU usage z/os v1r13 Migrated back in October, z/OS 1.11 to 1.13. Saw nothing of the like. For some reason I kinda remember another email on this subject sometime in the last 6 months or so. Please don't hold me this to this but an exit may have needed reassembled and I wanna say it was iefactrt that was not working right. You might be able to search the archives for it. Sorry best this feeble brain can do. On 1/7/2013 2:09 PM, Crabtree, Anne D wrote: > Since bringing up z/os v1r13 in late October, our CPU stays at 100%, > batch jobs are backing up and people are complaining. (Migrated from > z/os v1r11) > > I've looked for information on increased CPU for 1.13 but haven't found much. > I see now that there is a product you can download to predict increased CPU > when migrating, but, unfortunately, I didn't realize that until now. I can't > say for sure what kind of percentage the CPU was peaking at for 1.11, but I > know that we did not get these phone calls complaining about response time! > > I just wondered if anyone else experienced a significant increase in CPU > usage going to 1.13? > > Anne D. Crabtree > System Programmer > WV Office of Technology Data Center > 1900 Kanawha Blvd East > Bldg 6, Room B-110 > Charleston, WV 25305 > (304)558-5914 ext 58292 > (304)558-1441 fax > > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, send > email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- Brian W. France Systems Administrator (Mainframe) Pennsylvania State University Administrative Information Services - Infrastructure/SYSARC Rm 25 Shields Bldg., University Park, Pa. 16802 814-863-4739 b...@psu.edu "To make an apple pie from scratch, you must first invent the universe." Carl Sagan -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: increased CPU usage z/os v1r13
Migrated back in October, z/OS 1.11 to 1.13. Saw nothing of the like. For some reason I kinda remember another email on this subject sometime in the last 6 months or so. Please don't hold me this to this but an exit may have needed reassembled and I wanna say it was iefactrt that was not working right. You might be able to search the archives for it. Sorry best this feeble brain can do. On 1/7/2013 2:09 PM, Crabtree, Anne D wrote: Since bringing up z/os v1r13 in late October, our CPU stays at 100%, batch jobs are backing up and people are complaining. (Migrated from z/os v1r11) I've looked for information on increased CPU for 1.13 but haven't found much. I see now that there is a product you can download to predict increased CPU when migrating, but, unfortunately, I didn't realize that until now. I can't say for sure what kind of percentage the CPU was peaking at for 1.11, but I know that we did not get these phone calls complaining about response time! I just wondered if anyone else experienced a significant increase in CPU usage going to 1.13? Anne D. Crabtree System Programmer WV Office of Technology Data Center 1900 Kanawha Blvd East Bldg 6, Room B-110 Charleston, WV 25305 (304)558-5914 ext 58292 (304)558-1441 fax -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- Brian W. France Systems Administrator (Mainframe) Pennsylvania State University Administrative Information Services - Infrastructure/SYSARC Rm 25 Shields Bldg., University Park, Pa. 16802 814-863-4739 b...@psu.edu "To make an apple pie from scratch, you must first invent the universe." Carl Sagan -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
increased CPU usage z/os v1r13
Since bringing up z/os v1r13 in late October, our CPU stays at 100%, batch jobs are backing up and people are complaining. (Migrated from z/os v1r11) I've looked for information on increased CPU for 1.13 but haven't found much. I see now that there is a product you can download to predict increased CPU when migrating, but, unfortunately, I didn't realize that until now. I can't say for sure what kind of percentage the CPU was peaking at for 1.11, but I know that we did not get these phone calls complaining about response time! I just wondered if anyone else experienced a significant increase in CPU usage going to 1.13? Anne D. Crabtree System Programmer WV Office of Technology Data Center 1900 Kanawha Blvd East Bldg 6, Room B-110 Charleston, WV 25305 (304)558-5914 ext 58292 (304)558-1441 fax -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN