Re: Z196 Cobol pgm with higher CPU
Hi I didn't see What was the solution ? On 7/14/2011 7:32 PM, David Mingee wrote: Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Any help would be greatly appreciated. -- 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
Re: Z196 Cobol pgm with higher CPU
Hello, We still have no solution. We checked the MCL and it was good. -Original Message- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@bama.ua.edu] On Behalf Of Miklos Szigetvari Sent: Thursday, July 21, 2011 9:01 AM To: IBM-MAIN@bama.ua.edu Subject: Re: Z196 Cobol pgm with higher CPU Hi I didn't see What was the solution ? On 7/14/2011 7:32 PM, David Mingee wrote: Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Any help would be greatly appreciated. -- 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 -- 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
Re: Z196 Cobol pgm with higher CPU
Some other weird things to check. 1. Is the cpu increase in DB2? check the DB2 accounting records for this. 2. Someone mentioned Cache issues - that would require the program be compiled without RENT to get working storage in the same memory area - CBT has a program that will extract cobol compile options. 3. brute force method - get a STROBE report and look for where the most cpu is. I know not everyone has STROBE. Mike On Thu, Jul 21, 2011 at 8:57 AM, Mingee, David david.min...@libertymutual.com wrote: Hello, We still have no solution. We checked the MCL and it was good. -Original Message- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@bama.ua.edu] On Behalf Of Miklos Szigetvari Sent: Thursday, July 21, 2011 9:01 AM To: IBM-MAIN@bama.ua.edu Subject: Re: Z196 Cobol pgm with higher CPU Hi I didn't see What was the solution ? On 7/14/2011 7:32 PM, David Mingee wrote: Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Any help would be greatly appreciated. -- 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 -- 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 -- Mike -- 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
Re: Z196 Cobol pgm with higher CPU
Do you have the MCL/data stream number ? we have looked on resource link and haven't found anything -- 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
Re: Z196 Cobol pgm with higher CPU
and, the strobes that we have run show a COBOL move with an MVCL being generated in the PMAP. I have asked IBM for the datastream number -- 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
Re: Z196 Cobol pgm with higher CPU
On 7/15/2011 2:11 PM, J Ellis wrote: and, the strobes that we have run show a COBOL move with an MVCL being generated in the PMAP. I have asked IBM for the datastream number -- 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 Just to know, what was the previous CPU ? Maybe what is this datastream number ? -- 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
Re: Z196 Cobol pgm with higher CPU
I just had it checked and it is MCL 089, available in februari. Kees. J Ellis jerry.el...@libertymutual.com wrote in message news:6693309261138301.wa.jerry.ellislibertymutual@bama.ua.edu... Do you have the MCL/data stream number ? we have looked on resource link and haven't found anything -- 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 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 lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: Z196 Cobol pgm with higher CPU
I just had it checked and it is MCL 089, available in februari. RETAIN info: MCL fix U799O089 EC N29799 B28B. Desc: performance issue seen on customer applications which heavily use the move char long, move char long extended and move char long Unicode instructions. Kees. J Ellis jerry.el...@libertymutual.com wrote in message news:6693309261138301.wa.jerry.ellislibertymutual@bama.ua.edu... Do you have the MCL/data stream number ? we have looked on resource link and haven't found anything -- 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 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 lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: Z196 Cobol pgm with higher CPU
I have it on, next step will be to re-compile, thanks to all snip I just had it checked and it is MCL 089, available in februari. RETAIN info: MCL fix U799O089 EC N29799 B28B. Desc: performance issue seen on customer applications which heavily use the move char long, move char long extended and move char long Unicode instructions. Kees. -- 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
Re: Z196 Cobol pgm with higher CPU
May I surmise from the initials MCL that this is a micro- or milli-code fix for a z196 machine? Is it possible that such a fix has implications on lower architecture boxes like z10's? I have no access to IBMLINK or RETAIN or I would look these things up myself. The reason I ask such a question is that in several application-oriented CPU reduction efforts in recent years involving COBOL code, I found that some quite large reductions in average CPU time were gained from the elimination or at least from the minimization of MVCL and CLCL instructions in the generated code. If the z196 has a speedup fix for these instructions and if that fix could be back-ported to z10's or even to lower-order boxes, a possibly large amount of COBOL application CPU time could be saved over the z universe... Peter -Original Message- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@bama.ua.edu] On Behalf Of Vernooij, CP - SPLXM Sent: Friday, July 15, 2011 8:45 AM To: IBM-MAIN@bama.ua.edu Subject: Re: Z196 Cobol pgm with higher CPU I just had it checked and it is MCL 089, available in februari. RETAIN info: MCL fix U799O089 EC N29799 B28B. Desc: performance issue seen on customer applications which heavily use the move char long, move char long extended and move char long Unicode instructions. -- This message and any attachments are intended only for the use of the addressee and may contain information that is privileged and confidential. If the reader of the message is not the intended recipient or an authorized representative of the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by e-mail and delete the message and any attachments from your system. -- 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
Re: Z196 Cobol pgm with higher CPU
I'm not sure if the past few posts are in response to my comment about a crippling COBOL/MVCL problem. If so--and it was not my dog, as Schmuel would say--the offending application had previously worked OK on earlier z hardware. Moving from z10 to z196 was a day-one disaster. Once the smoking MVCL was implicated, the application was improved dramatically with a minor rewrite of just a few lines of source code. The z196 problem resulted from excessive MVCL execution in the compiled code. MCL was eventually updated in some way to avoid the problem. 'Reasonable' use of MVCL would not have caused a crisis in the first place. The application induced the problem, which did not raise any alarm on earlier hardware. . . JO.Skip Robinson SCE Infrastructure Technology Services Electric Dragon Team Paddler SHARE MVS Program Co-Manager 626-302-7535 Office 323-715-0595 Mobile jo.skip.robin...@sce.com From: Farley, Peter x23353 peter.far...@broadridge.com To: IBM-MAIN@bama.ua.edu Date: 07/15/2011 07:51 AM Subject:Re: Z196 Cobol pgm with higher CPU Sent by:IBM Mainframe Discussion List IBM-MAIN@bama.ua.edu May I surmise from the initials MCL that this is a micro- or milli-code fix for a z196 machine? Is it possible that such a fix has implications on lower architecture boxes like z10's? I have no access to IBMLINK or RETAIN or I would look these things up myself. The reason I ask such a question is that in several application-oriented CPU reduction efforts in recent years involving COBOL code, I found that some quite large reductions in average CPU time were gained from the elimination or at least from the minimization of MVCL and CLCL instructions in the generated code. If the z196 has a speedup fix for these instructions and if that fix could be back-ported to z10's or even to lower-order boxes, a possibly large amount of COBOL application CPU time could be saved over the z universe... Peter -Original Message- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@bama.ua.edu] On Behalf Of Vernooij, CP - SPLXM Sent: Friday, July 15, 2011 8:45 AM To: IBM-MAIN@bama.ua.edu Subject: Re: Z196 Cobol pgm with higher CPU I just had it checked and it is MCL 089, available in februari. RETAIN info: MCL fix U799O089 EC N29799 B28B. Desc: performance issue seen on customer applications which heavily use the move char long, move char long extended and move char long Unicode instructions. -- 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
Re: Z196 Cobol pgm with higher CPU
In of046b3d6f.6a422665-on882578ce.0055252c-882578ce.0055f...@sce.com, on 07/15/2011 at 08:38 AM, Skip Robinson jo.skip.robin...@sce.com said: 'Reasonable' use of MVCL would not have caused a crisis in the first place. What was unreasonable about the application's and compiler's use of MVCL? From what you described it was the z196 that was unreasonable. -- Shmuel (Seymour J.) Metz, SysProg and JOAT ISO position; see http://patriot.net/~shmuel/resume/brief.html We don't care. We don't have to care, we're Congress. (S877: The Shut up and Eat Your spam act of 2003) -- 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
Re: Z196 Cobol pgm with higher CPU
(I'm speaking on behalf of someone else's dog.) The code was badly written. Just how badly was never understood until the z196 choked on excessive MVCLs. Of course the hardware was at fault. That's why IBM supplied an MCL fix. But even before that fix was available, some relief was obtained from fixing sloppy code. BTW no one ever blamed the compiler AFAIK. . . JO.Skip Robinson SCE Infrastructure Technology Services Electric Dragon Team Paddler SHARE MVS Program Co-Manager 626-302-7535 Office 323-715-0595 Mobile jo.skip.robin...@sce.com From: Shmuel Metz (Seymour J.) shmuel+ibm-m...@patriot.net To: IBM-MAIN@bama.ua.edu Date: 07/15/2011 11:13 AM Subject:Re: Z196 Cobol pgm with higher CPU Sent by:IBM Mainframe Discussion List IBM-MAIN@bama.ua.edu In of046b3d6f.6a422665-on882578ce.0055252c-882578ce.0055f...@sce.com, on 07/15/2011 at 08:38 AM, Skip Robinson jo.skip.robin...@sce.com said: 'Reasonable' use of MVCL would not have caused a crisis in the first place. What was unreasonable about the application's and compiler's use of MVCL? From what you described it was the z196 that was unreasonable. -- 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
Z196 Cobol pgm with higher CPU
Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Any help would be greatly appreciated. -- 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
Re: Z196 Cobol pgm with higher CPU
My first thought is data proximity to instructions. That is data being modified within the width of the processor I-cache. The data modification will cause cache flush of the I-cache, and many processor cycles to re-establish the cache and various pipelines. This happened with SAS when the z/900 (?) processors came out. There should be a discussion in the archives about the SAS version of this issue. The solution is to move the data far enough away from the instructions so that the modification does not cause the cache to be flushed. HTH, snip Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Any help would be greatly appreciated. /snip -- 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
Re: Z196 Cobol pgm with higher CPU
just a guess but I expect the program does dynamic SQL. That is the only way you can get a different access path without a manual rebind. CPU speed is part of the DB2 optimization so you should expect some issues when you do rebinds and install new versions of programs. And there is more expertise on DB2 listserv db...@lists.idug.org Mike -- 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
Re: Z196 Cobol pgm with higher CPU
I know of a case where one specific COBOL application went to h*ll in a h*nd basket on z196 after performing satisfactorily for years on z10 and earlier. The problem turned out to be handling of MVCL, which this application was replete with. The problem was fixed in a microcode level that I would expect everyone to have by now. . . JO.Skip Robinson SCE Infrastructure Technology Services Electric Dragon Team Paddler SHARE MVS Program Co-Manager 626-302-7535 Office 323-715-0595 Mobile jo.skip.robin...@sce.com From: Staller, Allan allan.stal...@kbmg.com To: IBM-MAIN@bama.ua.edu Date: 07/14/2011 10:52 AM Subject:Re: Z196 Cobol pgm with higher CPU Sent by:IBM Mainframe Discussion List IBM-MAIN@bama.ua.edu My first thought is data proximity to instructions. That is data being modified within the width of the processor I-cache. The data modification will cause cache flush of the I-cache, and many processor cycles to re-establish the cache and various pipelines. This happened with SAS when the z/900 (?) processors came out. There should be a discussion in the archives about the SAS version of this issue. The solution is to move the data far enough away from the instructions so that the modification does not cause the cache to be flushed. HTH, snip Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Any help would be greatly appreciated. /snip -- 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
Re: Z196 Cobol pgm with higher CPU
So, since it is Cobol, would just a recompile with a recent compiler be sufficient to generate an efficient loadmodule? We discussed a similar issue internally recently with regards to the new instructions that are available on today's machines and were not 10 or 20 years ago when an old module was compiled. Also for this reason, recompiling it could make it more efficient. Kees. Staller, Allan allan.stal...@kbmg.com wrote in message news:45e5f2f45d7878458ee5ca679697335502e25...@usdaexch01.kbm1.loc... My first thought is data proximity to instructions. That is data being modified within the width of the processor I-cache. The data modification will cause cache flush of the I-cache, and many processor cycles to re-establish the cache and various pipelines. This happened with SAS when the z/900 (?) processors came out. There should be a discussion in the archives about the SAS version of this issue. The solution is to move the data far enough away from the instructions so that the modification does not cause the cache to be flushed. HTH, snip Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Any help would be greatly appreciated. /snip -- 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 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 lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: Z196 Cobol pgm with higher CPU
David Mingee wrote: Hello, We recently installed a Z196 and we have one COBOL DB2 program that is using twice the CPU time as before the new CPU. The program has not been recompiled and the data volume is the same. Did you do any z/OS and/or DB2 upgrade too? What versions are you having now when having this problem? IBM has found nothing, yet. Strobe has not revealed any answers. All other programs use about 1/2 of the CPU time now. Why not re-check that program? If you can recompile that program with the latest LE + DB2 libraries and still having CPU time problem, you need to sail away to PTF and APAR land using a PMR ... ;-D Groete / Greetings Elardus Engelbrecht Don't smoke in bed. And don't sleep in ashtrays! ;-D -- 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
Re: Z196 Cobol pgm with higher CPU
The last PMAP (or whatever it is called now) I looked at showed that the Cobol code generator was not even using relative branching. IBM Mainframe Discussion List IBM-MAIN@bama.ua.edu wrote on 07/14/2011 02:04:37 PM: From: Vernooij, CP - SPLXM kees.verno...@klm.com So, since it is Cobol, would just a recompile with a recent compiler be sufficient to generate an efficient loadmodule? We discussed a similar issue internally recently with regards to the new instructions that are available on today's machines and were not 10 or 20 years ago when an old module was compiled. Also for this reason, recompiling it could make it more efficient. - The information contained in this communication (including any attachments hereto) is confidential and is intended solely for the personal and confidential use of the individual or entity to whom it is addressed. If the reader of this message is not the intended recipient or an agent responsible for delivering it to the intended recipient, you are hereby notified that you have received this communication in error and that any review, dissemination, copying, or unauthorized use of this information, or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this communication in error, please notify us immediately by e-mail, and delete the original message. Thank you -- 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