Re: HCPPGT401I 90% of spool
That's a really good point, Rob. A thrashing system that overflows page space into spool can chew through spool at a far faster rate than even th e most egregious user / application spool abuse. (The paging subsystem has far higher bandwidth than the spool subsystem, both in terms of allocatio n and actual data movement.) Worse, paging using up spool space will almos t certainly lead to a PGT004 hard abend, whereas spool using up spool space normally just leads to spool files being closed and new writes lost. - Bill Holder z/VM Development, IBM
Re: HCPPGT401I 90% of spool
On Fri, Jul 11, 2008 at 5:10 PM, Mary Anne Matyaz [EMAIL PROTECTED] wrote: Question on paging though. You're supposed to have, like twice the amount of allocated guest memory in paging space. So assuming I do, will I ever fill it up? IE, if I total up the memory from user direct, double it, and have that much paging space (though I have tripled it to be safe), can I ever fill it up? Eventually things will fill up when you're not looking ;-) For example because someone was going to add the new page pack later and forgot. Or because you have given out fairly large VDISKs for swap that don't get used normally (you don't want to match twice the size of all unused VDISK space in paging, I think). On most systems there should be a lot of time to react when you have the alert at 50% (and once you hit that threshold, you look back in history to see how it got there and who did it, because you may have other options than adding page space). Rob -- Rob van der Heij Velocity Software http://velocitysoftware.com/
RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400
97400 Routine: Common Severity: ERROR Explanation: Sever condition returned from APPC/VM communication request. If your application receives this reason code intermittently, but the file pool is still available and other commands work correctly, the file pool server may be improperly configured. Notify your file pool administrator of this condition. (Ask the administrator to check the USERS start-up parameter value.) OK ... what should I be asking myself about Ask the administrator to check the USERS start-up parameter value.? This FilePool was just created from scratch via FILESERV GENERATE, so it's not a case where the USERS parameter may have been changed, but the server not rebuilt. I definitely get the error intermittently as the not-so-clear explanation states (it's an empty FilePool I'm populating from a backup to tape). The correlation between USERS and WHAT is not correct ... ? JR (Steven) Imler CA Senior Software Engineer Tel: +1 703 708 3479 Fax: +1 703 708 3267 [EMAIL PROTECTED]
HiperSockets
Hi Does anyone if there is a restriction that forces you to start the HiperSockets addresses on an even number of the Triplet? Ex: Could I start at EA03 (odd number) for 3 (EA03, EA04, EA05)? Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED]
Re: RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400
Steve, The formula for determining the USERS is not at all accurate. I once had a small filepool on a system that only had 35 userids in its directory. I found that with a value of 100 for USERS, which was higher than the formula suggested as only 16 of the users were even granted authority to access the single defined filespace, the system would rapidly grind to a halt. By upping the value for USERS to 1000, things ran smoothly. Regards, Richard Schuh -Original Message- From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of Imler, Steven J Sent: Friday, July 11, 2008 10:04 AM To: IBMVM@LISTSERV.UARK.EDU Subject: RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400 97400 Routine: Common Severity: ERROR Explanation: Sever condition returned from APPC/VM communication request. If your application receives this reason code intermittently, but the file pool is still available and other commands work correctly, the file pool server may be improperly configured. Notify your file pool administrator of this condition. (Ask the administrator to check the USERS start-up parameter value.) OK ... what should I be asking myself about Ask the administrator to check the USERS start-up parameter value.? This FilePool was just created from scratch via FILESERV GENERATE, so it's not a case where the USERS parameter may have been changed, but the server not rebuilt. I definitely get the error intermittently as the not-so-clear explanation states (it's an empty FilePool I'm populating from a backup to tape). The correlation between USERS and WHAT is not correct ... ? JR (Steven) Imler CA Senior Software Engineer Tel: +1 703 708 3479 Fax: +1 703 708 3267 [EMAIL PROTECTED]
Re: RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400
Hmm ... how embarrassing. I of all people should have known to check the SFS server console for oddness. snip DISCONNECT AT 09:42:44 EDT FRIDAY 07/11/08 DMS4GE3245I 07-11-08 12:06:50 The log is 91% full DMS4GE3245I 07-11-08 12:06:57 The log is 92% full DMS4GE3247W 07-11-08 12:07:04 Automatic LUW rollback processing initiated z/VM Version 5 Release 3.0, Service Level 0703 (64-bit), snip That being said, what's the correlation between the error I got and the LOG filling up? (At least I assume the LOG filled up?) JR JR (Steven) Imler CA Senior Software Engineer Tel: +1 703 708 3479 Fax: +1 703 708 3267 [EMAIL PROTECTED]
Re: HiperSockets
Thanks Brian! Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] -Original Message- From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of Brian Nielsen Sent: Friday, July 11, 2008 1:28 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: HiperSockets There is no current restriction. We use our hipersocket addresses in consecutve groups of 3 without regard to the evenness or oddness of the = real address. Brian Nielsen On Fri, 11 Jul 2008 13:10:58 -0400, Martin, Terry R. (CMS/CTR) (CTR) [EMAIL PROTECTED] wrote: Hi Does anyone if there is a restriction that forces you to start the HiperSockets addresses on an even number of the Triplet? Ex: Could I start at EA03 (odd number) for 3 (EA03, EA04, EA05)? Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED]
Re: HiperSockets
Hmm - Thought you could do that, as long as you used virtual addresses and the 1st address was even. Maybe I'm confused with another OS. I work on to many. On Fri, Jul 11, 2008 at 1:51 PM, Martin, Terry R. (CMS/CTR) (CTR) [EMAIL PROTECTED] wrote: Thanks Brian! Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] -Original Message- From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of Brian Nielsen Sent: Friday, July 11, 2008 1:28 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: HiperSockets There is no current restriction. We use our hipersocket addresses in consecutve groups of 3 without regard to the evenness or oddness of the = real address. Brian Nielsen On Fri, 11 Jul 2008 13:10:58 -0400, Martin, Terry R. (CMS/CTR) (CTR) [EMAIL PROTECTED] wrote: Hi Does anyone if there is a restriction that forces you to start the HiperSockets addresses on an even number of the Triplet? Ex: Could I start at EA03 (odd number) for 3 (EA03, EA04, EA05)? Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] -- Mark Pace Mainline Information Systems
Re: HiperSockets
Doesn't a CTC connection require the even/odd address for send/receive? -- John McKown Senior Systems Programmer HealthMarkets Keeping the Promise of Affordable Coverage Administrative Services Group Information Technology This message (including any attachments) contains confidential information intended for a specific individual and purpose, and its content is protected by law. If you are not the intended recipient, you should delete this message and are hereby notified that any disclosure, copying, or distribution of this transmission, or taking any action based on it, is strictly prohibited. From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of Mark Pace Sent: Friday, July 11, 2008 1:29 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: HiperSockets Hmm - Thought you could do that, as long as you used virtual addresses and the 1st address was even. Maybe I'm confused with another OS. I work on to many. [snip] -- Mark Pace Mainline Information Systems
Re: HiperSockets
Yes, CTC must be an even/odd pair. On Fri, Jul 11, 2008 at 2:30 PM, McKown, John [EMAIL PROTECTED] wrote: Doesn't a CTC connection require the even/odd address for send/receive? -- John McKown Senior Systems Programmer HealthMarkets Keeping the Promise of Affordable Coverage Administrative Services Group Information Technology This message (including any attachments) contains confidential information intended for a specific individual and purpose, and its content is protected by law. If you are not the intended recipient, you should delete this message and are hereby notified that any disclosure, copying, or distribution of this transmission, or taking any action based on it, is strictly prohibited. -- *From:* The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] *On Behalf Of *Mark Pace *Sent:* Friday, July 11, 2008 1:29 PM *To:* IBMVM@LISTSERV.UARK.EDU *Subject:* Re: HiperSockets Hmm - Thought you could do that, as long as you used virtual addresses and the 1st address was even. Maybe I'm confused with another OS. I work on to many. [snip] -- Mark Pace Mainline Information Systems -- Mark Pace Mainline Information Systems
Re: HiperSockets
Current versions of z/Linux don't care either. We have some that start a t even virtual addresses and some at odd virtual addresses. Older versions of Linux did have a restriction, I believe. This topic has been discusse d here before, so if you search you should be able to find them. Brian Nielsen On Fri, 11 Jul 2008 14:28:34 -0400, Mark Pace [EMAIL PROTECTED] wrote: Hmm - Thought you could do that, as long as you used virtual addresses a nd the 1st address was even. Maybe I'm confused with another OS. I work o n to many. On Fri, Jul 11, 2008 at 1:51 PM, Martin, Terry R. (CMS/CTR) (CTR) [EMAIL PROTECTED] wrote: Thanks Brian! Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] -Original Message- From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] O n Behalf Of Brian Nielsen Sent: Friday, July 11, 2008 1:28 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: HiperSockets There is no current restriction. We use our hipersocket addresses in consecutve groups of 3 without regard to the evenness or oddness of th e = real address. Brian Nielsen On Fri, 11 Jul 2008 13:10:58 -0400, Martin, Terry R. (CMS/CTR) (CTR) [EMAIL PROTECTED] wrote: Hi Does anyone if there is a restriction that forces you to start the HiperSockets addresses on an even number of the Triplet? Ex: Could I start at EA03 (odd number) for 3 (EA03, EA04, EA05)? Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] -- Mark Pace Mainline Information Systems
Re: HiperSockets
On Friday, 07/11/2008 at 02:32 EDT, McKown, John [EMAIL PROTECTED] wrote: Doesn't a CTC connection require the even/odd address for send/receive? No. It requires n and n+1. And even that is just an artifact of the device driver. There's nothing special about adjacent subchannels on a CTC. Your IOCP could even configure device number n and n+1 on different CTC chpids. It just doesn't matter... This even thing with OSAs and HiperSockets is the vestigal memory of the original OSA-1s that required the control path to be on an even address. I think that was lifted in the OSA-2 and was definitely gone by the time OSA-Express appeared. It ranks right up there with PORTNAMEs. They have been optional on z/VM and Linux for several years, yet I keep finding NEW configurations that have them coded and people saying they have to match, which they do, but ONLY IF YOU CODE THEM! Alan Altmark z/VM Development IBM Endicott
Re: RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400
On Friday, 07/11/2008 at 01:51 EDT, Imler, Steven J [EMAIL PROTECTED] wrote: That being said, what's the correlation between the error I got and the LOG filling up? (At least I assume the LOG filled up?) What was returned in the WUERROR field, if anything? Some error codes (i.e. implicit rollback) are part of the wuerror information. As to why a log that isn't actually full would cause a rollback, I don't know. Is there anything you can commit along the way? Alan Altmark z/VM Development IBM Endicott
Re: RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400
What I sometimes did in the past for such massive updates that make the log fill faster than log backup can clean: - place the updater in CP SLEEP - issue CP SEND VMSERVx BACKUP to empty the logs - unSLEEP the updater. 2008/7/11 Alan Altmark [EMAIL PROTECTED]: On Friday, 07/11/2008 at 01:51 EDT, Imler, Steven J [EMAIL PROTECTED] wrote: That being said, what's the correlation between the error I got and the LOG filling up? (At least I assume the LOG filled up?) What was returned in the WUERROR field, if anything? Some error codes (i.e. implicit rollback) are part of the wuerror information. As to why a log that isn't actually full would cause a rollback, I don't know. Is there anything you can commit along the way? Alan Altmark z/VM Development IBM Endicott -- Kris Buelens, IBM Belgium, VM customer support
Performance Tool Kit
Hi I am starting up the PERFSVM Machine and I am receiving the following error: FCXTCP576E Error number 13 for BIND call Does anyone know what this means. It is obviously causing me not to see most of the DATA fields in the MONITOR! Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED]
Re: Performance Tool Kit
Hi I take that back I seem to be getting my MONITOR information. It will still be interesting to know what it really meant. The error message does not offer a whole lot of information. Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] From: Martin, Terry R. (CMS/CTR) (CTR) Sent: Friday, July 11, 2008 5:08 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Performance Tool Kit Hi I am starting up the PERFSVM Machine and I am receiving the following error: FCXTCP576E Error number 13 for BIND call Does anyone know what this means. It is obviously causing me not to see most of the DATA fields in the MONITOR! Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED]
Re: Performance Tool Kit
I believe this means PERFSVM doesn't have permission to bind to whatever port it's trying to use. Make sure you have an entry in the PORT statement of the TCP/IP server's configuration file. Regards, Miguel Delapaz z/VM TCP/IP Development The IBM z/VM Operating System IBMVM@LISTSERV.UARK.EDU wrote on 07/11/2008 02:13:05 PM: Hi I take that back I seem to be getting my MONITOR information. It will still be interesting to know what it really meant. The error message does not offer a whole lot of information. Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] From: Martin, Terry R. (CMS/CTR) (CTR) Sent: Friday, July 11, 2008 5:08 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Performance Tool Kit Hi I am starting up the PERFSVM Machine and I am receiving the following error: FCXTCP576E Error number 13 for BIND call Does anyone know what this means. It is obviously causing me not to see most of the DATA fields in the MONITOR!
code maint [was: Re: REXX coding question.]
I checked with Glenn Vanderburg to be sure I was remembering the following story correctly. This comment by Chip Davis reminded me of it: On Thu, 10 Jul 2008, Chip Davis wrote: ... there comes a time when one must eschew elegance (and perhaps some efficiency) on behalf of the poor schmuck who will need to read, understand, and/or modify your code later. Especially if that poor schmuck might be yourself. David Lippke was an inspiration to me while at Texas AM. He and Pete Reynolds installed UTS on VM/HPO on one of the Amdahl's. I'll skip that story and ASCII and all that for now, but Lippke was exceptionally bright. (Remember Bitnet 1/2, the TCP/IP over RSCS project he did with Bruce Crabill? Stuff like that.) So ... one day Dave-o was eyeballing some code (probably C) and struggling to figure out what the author was trying to accomplish. This was apparently a really tricky statement. (PARSE? stem vars? bah!) The comment read roughly, If you don't understand what this does, then you shouldn't be reading my code!. Thereafter, he never again wrote such comments. (He was, in fact, the author of that code.) Funny, but eduficational. -- R;
Re: RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400
Don't know about WUERROR ... I'd have to repro and force a DUMP to find what/if there was a value. In the mean time, I got bigger LOG! That appears to have solved the problem. I'll run the test a few more times! JR JR (Steven) Imler CA Senior Software Engineer Tel: +1 703 708 3479 Fax: +1 703 708 3267 [EMAIL PROTECTED] -Original Message- From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of Alan Altmark Sent: Friday, July 11, 2008 04:23 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: RC = 12 FROM CSL FUNCTION DMSWRBLK; REASON: 97400 On Friday, 07/11/2008 at 01:51 EDT, Imler, Steven J [EMAIL PROTECTED] wrote: That being said, what's the correlation between the error I got and the LOG filling up? (At least I assume the LOG filled up?) What was returned in the WUERROR field, if anything? Some error codes (i.e. implicit rollback) are part of the wuerror information. As to why a log that isn't actually full would cause a rollback, I don't know. Is there anything you can commit along the way? Alan Altmark z/VM Development IBM Endicott
Re: Performance Tool Kit
Have you issued HELP FCX576E? 2008/7/11 Martin, Terry R. (CMS/CTR) (CTR) [EMAIL PROTECTED]: Hi I am starting up the PERFSVM Machine and I am receiving the following error: FCXTCP576E Error number 13 for BIND call Does anyone know what this means. It is obviously causing me not to see most of the DATA fields in the MONITOR! Thank You, Terry Martin Lockheed Martin - Information Technology z/OS z/VM Systems - Performance and Tuning Cell - 443 632-4191 Work - 410 786-0386 [EMAIL PROTECTED] -- Kris Buelens, IBM Belgium, VM customer support