Re: SYSPLEX for PDS-E Sharing
In [EMAIL PROTECTED], on 01/18/2007 at 12:48 PM, Rugen, Len [EMAIL PROTECTED] said: The alternative is to go over to the dark side.. Or to *bsd or Linux. -- 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 [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: SYSPLEX for PDS-E Sharing
In [EMAIL PROTECTED], on 01/18/2007 at 01:30 PM, Craddock, Chris [EMAIL PROTECTED] said: Don't forget that (AFAIK) all currently supported processors support some form of xMIF, EXPN? The last I heard, PR/SM still didn't provide any form of virtual CTCA. Or are you alluding to the fact that one pair of channels, cross connected[1], can serve as CTCA's for every pair of LPAR's? [1] Or connected to the same director/fabric. -- 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 [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: SYSPLEX for PDS-E Sharing
In [EMAIL PROTECTED], on 01/17/2007 at 01:42 PM, Rugen, Len [EMAIL PROTECTED] said: Do I need to setup GRS (yuck) in a ring to safely share PDSES? Star is preferred, but you do need GRS. I don't think I want or need XCF. XCF is part and parcel of sysplex. In [EMAIL PROTECTED], on 01/17/2007 at 03:39 PM, Rugen, Len [EMAIL PROTECTED] said: Isn't XCF the coupling facility? No; XCF can run over a CTCA for basic sysplex. I'm only a 2-way processor, wouldn't I have to sacrifice one processor to the CF LPAR? No. You don't need a CF for basic sysplex, and you can[1] share a process for the CF LPAR if you go that route. [1] There are, of course, performance ramifications. -- 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 [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: SYSPLEX for PDS-E Sharing
In [EMAIL PROTECTED], on 01/17/2007 at 05:45 PM, Jim Mulder [EMAIL PROTECTED] said: Basic sysplex came in MVS/ESA SP4.1. PDSE came in MVS/ESA SP4.3. PDSE came in as part of DFSMS. I don't recall which release. -- 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 [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: SYSPLEX for PDS-E Sharing
Craddock, Chris wrote: We started down the SYSPLEX road, but never could get hardware resources, then dropped from 6 to 3 LPAR's after Y2K. Setting up a ESCON CTC isn't all that difficult and won't cost you anything except a pair of ESCON channels. With that, you can set up a 3-LPAR SYSPLEX with GRS capabilities and go forward from there. Helps immensely if you diagram it out before you start to define the IOCDS entries. Don't forget that (AFAIK) all currently supported processors support some form of xMIF, meaning that you don't really need actual physical CTCs for connections within the same CEC/Box/your_favorite_name_for_a_box so if you're just connecting LPARS in the same machine you can do it trivially. soapbox Here we are in 2007. It is simply staggering to me that people are still whining about perceived problems and costs associated with sysplex. Those old chestnuts are bogus. There is NO GOOD REASON to run monoplexes in preference to at least a basic sysplex. There are. Different LPARs run different businesses, different companies. Different security rules. Strict for production, light for development. And if you really do need to run multiple LPARS for your work, then you need to do some soul searching about parallel sysplex. The benefits are enormous and the costs really aren't. I can't find *ANY* benefit of running *unrelated* systems on one CPC in sysplex. If you have more than one machine then pricing model could rape you to do sysplex, however it could be cheaper and more convenient to have *ONE* CPC. Sysplex always mean higher CPU utilization. Well tuned sysplex takes about 5% of CPU, but it can be 20-30% (it expected it during some tests). Base sysplex or even GRSplex is a cost. Relatively small in terms of investmens, however there's also software cost - laready mentioned higher CPU utilization. Parallel sysplex can improve work of sysplex, but the cost is significantly higher - ICF is approx. $120k. For single-CPC sysplex. When multi-CPC, you need much more equipment. We still talk about connecting unrelated systems into sysplex. Oh, one more benefit: migration effort. g Yes, I'm aware of RAS, five nines (99.999%) etc. However my business needs less nines, less costs. -- Radoslaw Skorupka Lodz, Poland -- 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
Security on development was Re: SYSPLEX for PDS-E Sharing
On 19 Jan 2007 04:37:48 -0800, in bit.listserv.ibm-main you wrote: Craddock, Chris wrote: We started down the SYSPLEX road, but never could get hardware resources, then dropped from 6 to 3 LPAR's after Y2K. Setting up a ESCON CTC isn't all that difficult and won't cost you anything except a pair of ESCON channels. With that, you can set up a 3-LPAR SYSPLEX with GRS capabilities and go forward from there. Helps immensely if you diagram it out before you start to define the IOCDS entries. Don't forget that (AFAIK) all currently supported processors support some form of xMIF, meaning that you don't really need actual physical CTCs for connections within the same CEC/Box/your_favorite_name_for_a_box so if you're just connecting LPARS in the same machine you can do it trivially. soapbox Here we are in 2007. It is simply staggering to me that people are still whining about perceived problems and costs associated with sysplex. Those old chestnuts are bogus. There is NO GOOD REASON to run monoplexes in preference to at least a basic sysplex. There are. Different LPARs run different businesses, different companies. Different security rules. Strict for production, light for development. Gjiven the source of most test data, I believe that we in the development area have been very lax in security. Being light on security in the development area will come back to haunt us. In most of my development work I have had access to what was basically some form of a copy of production data which has then been manipulated. rest snipped -- 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
Re: Security on development was Re: SYSPLEX for PDS-E Sharing
I said Those old chestnuts are bogus. There is NO GOOD REASON to run monoplexes in preference to at least a basic sysplex. RS said There are. Different LPARs run different businesses, different companies. Different security rules. Strict for production, light for development. If each monoplex is a different business then ok. But that's not usually the case. It is most often just multiple LPARs within the same business, being operated as if it is still 1985. z/OS systems are not especially reliable or manageable when run like that. The reliability of the base operating system and subsystems are fine, but the net when you add in applications and workloads that push systems into uncomfortable places in the design envelope is instability and erratic service at best. The sysplex architecture is there for a reason, not just to make IBM wealthy. Clark said Gjiven the source of most test data, I believe that we in the development area have been very lax in security. Being light on security in the development area will come back to haunt us. In most of my development work I have had access to what was basically some form of a copy of production data which has then been manipulated. True enough and this is part of the dirty little secret we live with. Security is only as good as the weakest link. Most systems run a mix of IBM, ISV and home-grown software and it is almost laughably easy (for a guy like me anyway) to find holes in that environment. It is entirely possible to make the platform all-but bulletproof, but most aren't. That's one of the reasons you hear me harp on about integrity from time to time. If you have integrity exposures, no matter how slight, you don't really have security. CC -- 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
Re: Security on development was Re: SYSPLEX for PDS-E Sharing
The previous shop I worked at had an MP3000-H50. Are you saying that instead of 1 Prod Lpar and 1 Test Lpar, we should have had 2 Prod Lpars and 2 Test Lpars so we could run in sysplex mode? We also would need to define a coupling facility, which I believe IBM always recommends 1 whole engine for. Of course, our H50 only had 1 engine. As I see it, a sysplex would not be very practical in that situation. Eric Bielefeld Sr. z/OS Systems Programmer Lands End Dodgeville, Wisconsin 414-475-7434 - Original Message - From: Craddock, Chris [EMAIL PROTECTED] I said Those old chestnuts are bogus. There is NO GOOD REASON to run monoplexes in preference to at least a basic sysplex. RS said There are. Different LPARs run different businesses, different companies. Different security rules. Strict for production, light for development. If each monoplex is a different business then ok. But that's not usually the case. It is most often just multiple LPARs within the same business, being operated as if it is still 1985. z/OS systems are not especially reliable or manageable when run like that. The reliability of the base operating system and subsystems are fine, but the net when you add in applications and workloads that push systems into uncomfortable places in the design envelope is instability and erratic service at best. The sysplex architecture is there for a reason, not just to make IBM wealthy. .html -- 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
Re: SYSPLEX for PDS-E Sharing
On Wed, 17 Jan 2007 17:45:43 -0500, Jim Mulder [EMAIL PROTECTED] wrote: Basic sysplex came in MVS/ESA SP4.1. PDSE came in MVS/ESA SP4.3. Jim Mulder z/OS System Test IBM Corp. Poughkeepsie, NY Thanks Jim. Brain cramp ... I don't know why I thought PDSE was around since ESA V3. Mark -- Mark Zelden Sr. Software and Systems Architect - z/OS Team Lead Zurich North America / Farmers Insurance Group - GITO mailto:[EMAIL PROTECTED] z/OS and OS390 expert at http://searchDataCenter.com/ateExperts/ Systems Programming expert at http://expertanswercenter.techtarget.com/ Mark's MVS Utilities: http://home.flash.net/~mzelden/mvsutil.html -- 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
Re: SYSPLEX for PDS-E Sharing
On Wed, 17 Jan 2007 17:11:44 -0500, Jack Kelly [EMAIL PROTECTED] wrote: if you have mim, why would you put grs on top of that? Because MIM doesn't protect PDSE. Search the archives. Don't forget IBM-MAIN-ARCHIVES. This for example: http://bama.ua.edu/cgi-bin/wa?A2=ind0309L=ibm-mainP=R25353 -- Tom Marchant -- 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
Re: SYSPLEX for PDS-E Sharing
We started down the SYSPLEX road, but never could get hardware resources, then dropped from 6 to 3 LPAR's after Y2K. Setting up a ESCON CTC isn't all that difficult and won't cost you anything except a pair of ESCON channels. With that, you can set up a 3-LPAR SYSPLEX with GRS capabilities and go forward from there. Helps immensely if you diagram it out before you start to define the IOCDS entries. Don't forget that (AFAIK) all currently supported processors support some form of xMIF, meaning that you don't really need actual physical CTCs for connections within the same CEC/Box/your_favorite_name_for_a_box so if you're just connecting LPARS in the same machine you can do it trivially. soapbox Here we are in 2007. It is simply staggering to me that people are still whining about perceived problems and costs associated with sysplex. Those old chestnuts are bogus. There is NO GOOD REASON to run monoplexes in preference to at least a basic sysplex. And if you really do need to run multiple LPARS for your work, then you need to do some soul searching about parallel sysplex. The benefits are enormous and the costs really aren't. /soapbox PDSE sharing requires all of the sharing members to be in the SAME sysplex. Period. PDSE (and a number of other components) share control information over XCF, which by definition restricts their sharing scope to members within the boundaries of the same sysplex. So even though you can share physical DASD resources across sysplex boundaries with a GRS ring or with MIM, you cannot share those (logical) datasets, no matter how you think it ought to work. CC -- 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
Re: SYSPLEX for PDS-E Sharing
This is about the 4th time I've installed the LAST OS upgrade we will ever need on the mainframe. Each one has been the max allowed upgrade just weeks prior to EOS of the current one. Many of us live under political rules, not technical ones. The alternative is to go over to the dark side.. soapbox Here we are in 2007. It is simply staggering to me that people are still whining about perceived problems and costs associated with sysplex. Those old chestnuts are bogus. There is NO GOOD REASON to run monoplexes in preference to at least a basic sysplex. And if you really do need to run multiple LPARS for your work, then you need to do some soul searching about parallel sysplex. The benefits are enormous and the costs really aren't. /soapbox -- 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
Re: SYSPLEX for PDS-E Sharing
On Thu, 18 Jan 2007 12:48:28 -0600, Rugen, Len wrote: This is about the 4th time I've installed the LAST OS upgrade we will ever need on the mainframe. Each one has been the max allowed upgrade just weeks prior to EOS of the current one. Many of us live under political rules, not technical ones. If you can't get your management to let you implement even a basic Sysplex, then any attempt to share PDSEs will be at your own peril. soapbox ... There is NO GOOD REASON to run monoplexes in preference to at least a basic sysplex. /soapbox -- Tom Marchant -- 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
Re: SYSPLEX for PDS-E Sharing
To share PDSEs, all LPARs must be a member of the same sysplex. From z/OS DFSMS: Using Data Sets, topic 3.8.9.3.3 Choosing Volumes for PDSEs in a Sysplex: -=-=-=-=- PDSEs are designed to be shared within a sysplex. When choosing volumes for PDSEs in a sysplex, be sure to follow these rules: o The volume serials for volumes that contain PDSEs must be unique within a sysplex. o A volume that contains PDSEs must not be open from more than one GRS complex at a time. o If PDSE extended sharing is active, a volume that contains a PDSE cannot be accessed from more than one sysplex at a time. In this context, a sysplex is all systems that can connect in a single XCF group, and a GRS complex is all the systems in a GRS configuration. A sysplex never spans more than one GRS complex. Note: for extended sharing, a PDSE can only be shared by the members of a GRS complex that are also members of the same sysplex. For example: in a six-system GRS complex, with four of the systems within the sysplex and two which are not, PDSEs can have extended sharing between the four members of the sysplex, but not the other two, non-sysplex systems. See z/OS MVS Planning: Global Resource Serialization for more information about the configurations that make up a GRS complex. If these volume assignment rules are not followed for PDSEs in a sysplex, data set accessibility or integrity may be impacted. -=-=-=-=- Brian On Wed, 17 Jan 2007 13:42:48 -0600, Rugen, Len wrote: I had hoped that our mainframe would be gone before I needed to exploit PDSEs for end users. Apparently Oracle 10G produces load modules that require LIBRARY datasets. We have 3 LPAR's on a z800 but not a SYSPLEX. Do I need to setup GRS (yuck) in a ring to safely share PDSES? The only other place I used PDSE's was for software install and they are read-only after they are created. How much pain is there to setup a GRS ring? When I RTFM, everything seems to talk about a XCF. I don't think I want or need XCF. -- 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
Re: SYSPLEX for PDS-E Sharing
For normal PDSE sharing, at the data set level only, you will need GRS, but not necessarily XCF. Check out the IBM Redbook PDSE Usage Guide for details. Don Imbriale -Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Rugen, Len Sent: Wednesday, January 17, 2007 2:43 PM To: IBM-MAIN@BAMA.UA.EDU Subject: SYSPLEX for PDS-E Sharing I had hoped that our mainframe would be gone before I needed to exploit PDSEs for end users. Apparently Oracle 10G produces load modules that require LIBRARY datasets. We have 3 LPAR's on a z800 but not a SYSPLEX. Do I need to setup GRS (yuck) in a ring to safely share PDSES? The only other place I used PDSE's was for software install and they are read-only after they are created. How much pain is there to setup a GRS ring? When I RTFM, everything seems to talk about a XCF. I don't think I want or need XCF. *** Bear Stearns is not responsible for any recommendation, solicitation, offer or agreement or any information about any transaction, customer account or account activity contained in this communication. *** -- 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
Re: SYSPLEX for PDS-E Sharing
On Wed, 17 Jan 2007 13:42:48 -0600, Len Rugen wrote: I had hoped that our mainframe would be gone before I needed to exploit PDSEs for end users. Apparently Oracle 10G produces load modules that require LIBRARY datasets. We have 3 LPAR's on a z800 but not a SYSPLEX. It is NOT safe to share a PDSE outside of a SYSPLEX. This has been discussed several times here. When you search the archives, make sure you also search the archives for IBM-MAIN-ARCHIVES. All of the older stuff is there. -- Tom Marchant -- 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
Re: SYSPLEX for PDS-E Sharing
On Wed, 2007-01-17 at 13:42 -0600, Rugen, Len wrote: How much pain is there to setup a GRS ring? When I RTFM, everything seems to talk about a XCF. I don't think I want or need XCF. Depends on how your systems are configured. If you have multiple systems all the same it can be a pain. Especially if you haven't used sysname (or similar) in dataset names. Can get *very* ugly when you start introducing systems that all have all their system datasets the same name, but on different volumes. Start with the default member in the books, and look for obvious candidates to add - but be ready for problems initially. Here's a vote for XCF - if you haven't done GRS setup before, it'll probably be *much* easier to get your head around XCF rather than GRS CTC. Shane ... -- 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
Re: SYSPLEX for PDS-E Sharing
I've probably missed something but I've found that setting up GRS via escon ctc (or whatever) has been rather straight forward. Most confusing thing has been assigning the right device numbers between the lpars/cpc. The sysname issue has always been more related to going to sysplex. The issue of same dsn on different volser is an issue but manageable. Jack Kelly LA Systems @ US Courts x 202-502-2390 -- 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
Re: SYSPLEX for PDS-E Sharing
Isn't XCF the coupling facility? I'm only a 2-way processor, wouldn't I have to sacrifice one processor to the CF LPAR? I was seeing all the crap about msys for setup and JVM's. Now that I read more, I'll just do it, it looks easier. I have CTC's for VTAM and MIM and extras that I think I can use for a GRS ring. I think it would be best to promote my production LPAR from a MONOPLEX to a SYSPLEX, then join my test LPARs. That should keep me from breaking my CICS LOGGER datasets in production. We started down the SYSPLEX road, but never could get hardware resources, then dropped from 6 to 3 LPAR's after Y2K. The mainframe has been going away for many years here also, now it's within 3 years, down from 5 years in 1998. My knowledge and training is appropriately stagnant as well. Thanks -Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Shane Sent: Wednesday, January 17, 2007 3:20 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: SYSPLEX for PDS-E Sharing On Wed, 2007-01-17 at 13:42 -0600, Rugen, Len wrote: How much pain is there to setup a GRS ring? When I RTFM, everything seems to talk about a XCF. I don't think I want or need XCF. Depends on how your systems are configured. If you have multiple systems all the same it can be a pain. Especially if you haven't used sysname (or similar) in dataset names. Can get *very* ugly when you start introducing systems that all have all their system datasets the same name, but on different volumes. Start with the default member in the books, and look for obvious candidates to add - but be ready for problems initially. Here's a vote for XCF - if you haven't done GRS setup before, it'll probably be *much* easier to get your head around XCF rather than GRS CTC. Shane ... -- 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 -- 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
Re: SYSPLEX for PDS-E Sharing
XCF is the cross-system communication facility. No, you won't have to sacrifice a CP. Jon snip Isn't XCF the coupling facility? I'm only a 2-way processor, wouldn't I have to sacrifice one processor to the CF LPAR? /snip -- 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
Re: SYSPLEX for PDS-E Sharing
I have CTC's for VTAM and MIM and extras that I think I can use for a GRS ring. Oh! You didn't mention you had MIM! I don't remember the details, but we did it accross two SYSPLEX environments with MIM almost 10 years ago. I'm no longer at that company, and it was out-sourced in 2001, so I can't get the details. But, it was doable in June 1999. We did it as part of our Goal Mode/TSO Generic Resource implementation, which I led. We had a roll-your-own change management application, that we converted the doc PDS to PDSE, and then we implemented the application in another SYSPLEX that was in our MIMPLEX. . Questions? Concerns? (Screems of Outrage?) -- 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
Re: SYSPLEX for PDS-E Sharing
On Wed, 17 Jan 2007 14:56:56 -0600, Tom Marchant [EMAIL PROTECTED] wrote: It is NOT safe to share a PDSE outside of a SYSPLEX. This has been discussed several times here. When you search the archives, make sure you also search the archives for IBM-MAIN-ARCHIVES. All of the older stuff is there. I haven't really thought about this before... PDSE has been around since ESA V3. Basic sysplex came in ESA 4.2 (right?). So how was PDSE sharing done prior to XCF - even the equivalent of what is now called PDSESHARING(NORMAL)? -- Mark Zelden Sr. Software and Systems Architect - z/OS Team Lead Zurich North America / Farmers Insurance Group - GITO mailto:[EMAIL PROTECTED] z/OS and OS390 expert at http://searchDataCenter.com/ateExperts/ Systems Programming expert at http://expertanswercenter.techtarget.com/ Mark's MVS Utilities: http://home.flash.net/~mzelden/mvsutil.html -- 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
Re: SYSPLEX for PDS-E Sharing
IBM Mainframe Discussion List IBM-MAIN@BAMA.UA.EDU wrote on 01/17/2007 05:35:15 PM: On Wed, 17 Jan 2007 14:56:56 -0600, Tom Marchant [EMAIL PROTECTED] wrote: It is NOT safe to share a PDSE outside of a SYSPLEX. This has been discussed several times here. When you search the archives, make sure you also search the archives for IBM-MAIN-ARCHIVES. All of the older stuff is there. I haven't really thought about this before... PDSE has been around since ESA V3. Basic sysplex came in ESA 4.2 (right?). So how was PDSE sharing done prior to XCF - even the equivalent of what is now called PDSESHARING(NORMAL)? Basic sysplex came in MVS/ESA SP4.1. PDSE came in MVS/ESA SP4.3. Jim Mulder z/OS System Test IBM Corp. Poughkeepsie, NY -- 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
Re: SYSPLEX for PDS-E Sharing
snip-- Isn't XCF the coupling facility? I'm only a 2-way processor, wouldn't I have to sacrifice one processor to the CF LPAR? I was seeing all the crap about msys for setup and JVM's. Now that I read more, I'll just do it, it looks easier. I have CTC's for VTAM and MIM and extras that I think I can use for a GRS ring. I think it would be best to promote my production LPAR from a MONOPLEX to a SYSPLEX, then join my test LPARs. That should keep me from breaking my CICS LOGGER datasets in production. We started down the SYSPLEX road, but never could get hardware resources, then dropped from 6 to 3 LPAR's after Y2K. The mainframe has been going away for many years here also, now it's within 3 years, down from 5 years in 1998. My knowledge and training is appropriately stagnant as well. -unsnip- Setting up a ESCON CTC isn't all that difficult and won't cost you anything except a pair of ESCON channels. With that, you can set up a 3-LPAR SYSPLEX with GRS capabilities and go forward from there. Helps immensely if you diagram it out before you start to define the IOCDS entries. -- 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