Re: z/OS 1.6 and more than 16 processors
Supposedly not. However, I can imagine some vendor of Java-based middleware doing so in the future. And likely charging the zAAP site for the total number of CPUs and zAAPs. Hum, I wonder if Oracle will do this. Some of their stuff is Java based, isn't it? Oracle uses their own JVM in the database server, so the zAAP will not be used. Oracle charges by the total number of engines that can execute Oracle. This is the number of Logical CPs in the Logical Partition. Their JVM is only used for stored procedures written in Java. The Oracle Installer (OUI) is a Java based GUI that installs Oracle products. This uses the host JVM, and so would use a zAAP if there was one available, but this has no effect on normal operation. Tom Russell Stay calm. Be brave. Wait for the signs. -- Jasper FriendlyBear -- 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
z/OS 1.6 and more than 16 processors
I intend to put z/OS 1.6 on a T-REX 2084-318 (18 processors). Has someone used this kind of configuration, z/OS 1.6 and more than 16 processors? What about performance? Any problem with ISVs products? Maria Inês H. S. Silva BANCO ITAU S/A São Paulo - Brasil Tel.: (0xx11) 3274-9087 E-mail: [EMAIL PROTECTED] Esta mensagem e uma correspondencia reservada. Se voce a recebeu por engano, por favor desconsidere-a. O sistema de mensagens da Internet nao e considerado seguro ou livre de erros. Esta instituicao nao se responsabiliza por opinioes ou declaracoes veiculadas atraves de e-mails. -- 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: z/OS 1.6 and more than 16 processors
On Thu, 7 Jul 2005 09:13:35 -0300, maria Ines maria- [EMAIL PROTECTED] wrote: I intend to put z/OS 1.6 on a T-REX 2084-318 (18 processors). Has someone used this kind of configuration, z/OS 1.6 and more than 16 processors? What about performance? Any problem with ISVs products? I don't have a single LPAR with more than 16 (yet?), but a friend of mine is running that configuration at his shop. I don't think there were any specific ISV issues because of 16 CPs. Obviously monitors need to support 16 to report correctly, but even if your monitor wasn't current enough on maintenance to support it, I doubt it would cause you a problem. Assuming you updated your ISV software for z/OS 1.6 and your ISV software vendors say the levels you are running are z/OS 1.6 ready, then they should be ready to support 16 CPs also. Regards, Mark -- Mark Zelden Sr. Software and Systems Architect mailto: [EMAIL PROTECTED] Systems Programming expert at http://Search390.com/ateExperts/ 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: z/OS 1.6 and more than 16 processors
Martin Packer wrote: Maria, will the z/OS R.6 LPAR have more than 16 logical CPs? If not I know of several of my customers who have machines with more than 16 Physical CPs - but all of those have no more than 16 Logical CPs in a partition. (I think it's just a matter of time before the z/OS R.6 support gets exploited by customers.) It's probably worth saying that support was up to 24 processors (the sum of CPs and zAAPs) in a single z/OS image at GA, and is now up to 32 processors in a single image. snip -- John Eells z/OS Technical Marketing IBM Poughkeepsie [EMAIL PROTECTED] -- 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: z/OS 1.6 and more than 16 processors
In a message dated 7/7/2005 10:39:48 A.M. Central Standard Time, [EMAIL PROTECTED] writes: It's probably worth saying that support was up to 24 processors (the sum of CPs and zAAPs) in a single z/OS image at GA, and is now up to 32 processors in a single image. Probably wouldn't hurt to review D/R scenarios too. -- 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: z/OS 1.6 and more than 16 processors
John, I didn't think 32 engine support was available until 1.7. Are you saying it's in 1.6 (with the right PTFs)? Jim Horne Lowe's Companies, Inc. John Eells wrote: It's probably worth saying that support was up to 24 processors (the sum of CPs and zAAPs) in a single z/OS image at GA, and is now up to 32 processors in a single image. -- 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: z/OS 1.6 and more than 16 processors
In a message dated 7/7/2005 10:39:48 A.M. Central Daylight Time, [EMAIL PROTECTED] writes: It's probably worth saying that support was up to 24 processors (the sum of CPs and zAAPs) in a single z/OS image at GA, and is now up to 32 processors in a single image. Obviously I missed some processor announcements. I know what an Instruction Processor and an I/O Processor are, but what is a zAAP? In what IBM book might I learn more about configurations that have 16 processors? I cannot find AAP in the z/Arch PoO book. And I am very curious as to how IBM managed to make z/OS work with more than 16 Instruction Processors in one image. E.g., how is the 17th CPU supported with the 16-bit CPU affinity mask in the SRB and all the other various 16-bit CPU masks imbedded at least in the Control Program? Bill Fairchild -- 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: z/OS 1.6 and more than 16 processors
On Thu, 7 Jul 2005 11:25:19 -0500, Paul Dineen [EMAIL PROTECTED] wrote: Bill, A zAAP (zSeries Application Assist Processor) is a processor engine which absorbs a percentage of JAVA workload, available for z990/z890. A large benefit of a zAAP is that OS software charges don't apply to it's use. http://www-1.ibm.com/servers/eserver/zseries/zaap/ And in case that doesn't sound familiar, you may have heard it referred to as an IFA (Integrated Facility for Applications - similar to Linux IFL). That was the name before the marketing people at IBM got involved. You still see that name used in various places like RMF reports and the HMC. Cheers, Mark -- Mark Zelden Sr. Software and Systems Architect mailto: [EMAIL PROTECTED] Systems Programming expert at http://Search390.com/ateExperts/ 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: z/OS 1.6 and more than 16 processors
Mark Zelden wrote: I don't have a single LPAR with more than 16 (yet?), but a friend of mine is running that configuration at his shop. I don't think there were any specific ISV issues because of 16 CPs. Obviously monitors need to support 16 to report correctly, but even if your monitor wasn't current enough on maintenance to support it, I doubt it would cause you a problem. Assuming you updated your ISV software for z/OS 1.6 and your ISV software vendors say the levels you are running are z/OS 1.6 ready, then they should be ready to support 16 CPs also. I would recommend making sure *all* ISV software is certified to run with 16 CPs before using such a configuration. I shudder to think about the potential for problems caused by an authorized/privileged state program loading a large (i.e., 15) CVTMAXMP value and looping through and/or updating control blocks and/or tables implicitly sized to 16 entries. Yikes! -- .-. | Edward E. Jaffe|| | Mgr, Research Development| [EMAIL PROTECTED]| | Phoenix Software International | Tel: (310) 338-0400 x318 | | 5200 W Century Blvd, Suite 800 | Fax: (310) 338-0801| | Los Angeles, CA 90045 | http://www.phoenixsoftware.com | '-' -- 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: z/OS 1.6 and more than 16 processors
-Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Jim Mulder Sent: Thursday, July 07, 2005 12:57 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: z/OS 1.6 and more than 16 processors z/OS does not allow SRB or TCB affinity to any processor whose address is larger than x'0F'. The affinity 16-bit masks were not extended. Some of the other 16-bit masks (in IHACSD, for example) were replaced by new (longer) bit masks. Jim Mulder z/OS System Test IBM Corp. Poughkeepsie, NY As a side question. Is there any particular use for CPU Affinity any more? IIRC, in the past, the CPUs could be asymmetrically configured (such as a CPU with Vector ability). Or even a CPU versus an AP (for those who remember the 158AP and 168AP, along with any other such animal). -- John McKown Senior Systems Programmer UICI Insurance Center 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. -- 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: z/OS 1.6 and more than 16 processors
In a message dated 7/7/2005 12:56:48 P.M. Central Daylight Time, [EMAIL PROTECTED] writes: z/OS does not allow SRB or TCB affinity to any processor whose address is larger than x'0F'. The affinity 16-bit masks were not extended. Some of the other 16-bit masks (in IHACSD, for example) were replaced by new (longer) bit masks. Awesome. I missed many announcements. Thanks for the info. Now I'm curious where the new knee of the curve is at which adding one additi onal instruction processor results in a net loss of throughput for a processor complex due to inter-processor serialization. Bill Fairchild -- 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: z/OS 1.6 and more than 16 processors
On Jul 7, 2005, at 11:25 AM, Paul Dineen wrote: Bill, A zAAP (zSeries Application Assist Processor) is a processor engine which absorbs a percentage of JAVA workload, available for z990/z890. A large benefit of a zAAP is that OS software charges don't apply to it's use. http://www-1.ibm.com/servers/eserver/zseries/zaap/ Paul Paul, Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? Ed -- 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: z/OS 1.6 and more than 16 processors
On Thu, 7 Jul 2005 12:59:39 -0500, McKown, John [EMAIL PROTECTED] wrote: As a side question. Is there any particular use for CPU Affinity any more? snip There have been a couple of farily recent threads about this. I think one a few months ago. Search the archives or google. Mark -- Mark Zelden Sr. Software and Systems Architect mailto: [EMAIL PROTECTED] Systems Programming expert at http://Search390.com/ateExperts/ 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: z/OS 1.6 and more than 16 processors
-Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Ed Gould Sent: Thursday, July 07, 2005 1:27 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: z/OS 1.6 and more than 16 processors snip Paul, Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? Ed IANP (I Am Not Paul), Supposedly not. However, I can imagine some vendor of Java-based middleware doing so in the future. And likely charging the zAAP site for the total number of CPUs and zAAPs. Hum, I wonder if Oracle will do this. Some of their stuff is Java based, isn't it? -- John McKown Senior Systems Programmer UICI Insurance Center 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. -- 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: z/OS 1.6 and more than 16 processors
On Thu, 7 Jul 2005 13:27:12 -0500, Ed Gould [EMAIL PROTECTED] wrote: Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? NO! At least not yet. That is the whole point. Well, almost the whole point. A zAAP engine is (IIRC)also about 1/4 the price of a general purpose processor. Mark -- Mark Zelden Sr. Software and Systems Architect mailto: [EMAIL PROTECTED] Systems Programming expert at http://Search390.com/ateExperts/ 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: z/OS 1.6 and more than 16 processors
On Jul 7, 2005, at 1:43 PM, McKown, John wrote: IANP (I Am Not Paul), Sorry, John. I made an error. Please accept my apologies. Ed Supposedly not. However, I can imagine some vendor of Java-based middleware doing so in the future. And likely charging the zAAP site for the total number of CPUs and zAAPs. Hum, I wonder if Oracle will do this. Some of their stuff is Java based, isn't it? -- John McKown Senior Systems Programmer UICI Insurance Center 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. -- 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: z/OS 1.6 and more than 16 processors
On Jul 7, 2005, at 1:44 PM, Mark Zelden wrote: On Thu, 7 Jul 2005 13:27:12 -0500, Ed Gould [EMAIL PROTECTED] wrote: Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? NO! At least not yet. That is the whole point. Well, almost the whole point. A zAAP engine is (IIRC)also about 1/4 the price of a general purpose processor. Mark Mark, Care to bet who will be the first? Ed -- 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: z/OS 1.6 and more than 16 processors
Mark Zelden wrote: On Thu, 7 Jul 2005 11:25:19 -0500, Paul Dineen [EMAIL PROTECTED] wrote: Bill, A zAAP (zSeries Application Assist Processor) is a processor engine which absorbs a percentage of JAVA workload, available for z990/z890. A large benefit of a zAAP is that OS software charges don't apply to it's use. http://www-1.ibm.com/servers/eserver/zseries/zaap/ And in case that doesn't sound familiar, you may have heard it referred to as an IFA (Integrated Facility for Applications - similar to Linux IFL). That was the name before the marketing people at IBM got involved. You still see that name used in various places like RMF reports and the HMC. IMHO it is also worth to say about LPAR weights with zAAP aka IFA. I was told (it's not my observation) about some gotcha's when different types of processors (CP, IFC, IFL, IFA) are in use, especially when CF LPARs are in use (due to 'active wait'). -- 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
Re: z/OS 1.6 and more than 16 processors
On Thu, 7 Jul 2005 13:43:48 -0500, McKown, John wrote: -Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Ed Gould Sent: Thursday, July 07, 2005 1:27 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: z/OS 1.6 and more than 16 processors snip Paul, Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? Ed IANP (I Am Not Paul), Supposedly not. However, I can imagine some vendor of Java-based middleware doing so in the future. And likely charging the zAAP site for the total number of CPUs and zAAPs. Hum, I wonder if Oracle will do this. Some of their stuff is Java based, isn't it? I'm waiting for the first license agreement that simply demands a seat on the company's board of directors and stock options for the vendor. We can't be all that far away from that, can we? -- Tom Schmidt Madison, WI -- 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: z/OS 1.6 and more than 16 processors
Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? Ed Ed, IBM does recommend checking with your vendor, but there are no software charges as zAAP engine is JAVA specific. Now if you're vendor software is JAVA based, I'd guess you may need to pay for zAAP capacity. The hardware cost is $125/K per zAAP engine. Paul -- 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: z/OS 1.6 and more than 16 processors
-Original Message- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Tom Schmidt Sent: Thursday, July 07, 2005 2:12 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: z/OS 1.6 and more than 16 processors On Thu, 7 Jul 2005 13:43:48 -0500, McKown, John wrote: snip I'm waiting for the first license agreement that simply demands a seat on the company's board of directors and stock options for the vendor. We can't be all that far away from that, can we? -- Tom Schmidt Well, if they are on the board, won't they be SOX candidates? Not likely that a vendor would want that. Stock options may not be all that attractive, depending on the licensee's stock price and history. -- John McKown Senior Systems Programmer UICI Insurance Center 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. -- 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: z/OS 1.6 and more than 16 processors
... Now I'm curious where the new knee of the curve is at which adding one additional instruction processor results in a net loss of throughput for a processor complex due to inter-processor serialization. ... It depends on who you ask. The LSPR throughput figures for anything above 16 are suspect. IBM doesn't directly admit it, but there is a one line disclaimer on the z/990 chart stating that everything above 16 is straight-lined. They didn't run anything above 16. There was a discussion last year on this. Take the numbers an plot them: there is no degradation after 16. -teD In God we Trust! All others bring data! --Deming -- 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: z/OS 1.6 and more than 16 processors
Horne, Jim - James S wrote: John, I didn't think 32 engine support was available until 1.7. Are you saying it's in 1.6 (with the right PTFs)? Yes. We announced the intent to support 32-way images on R6 in February: Scale up and scale out: z/OS V1.6 currently allows you to scale up in a single logical partition from 1 processor to 24 with good scalability and to scale out in a Parallel Sysplex® for higher availability. IBM plans to support up to 32 processors in a single logical partition on z990 servers on z/OS V1.6 in June 2005. Scalability is expected to be good all the way from 1 processor to 32. Notes * This is based on internal IBM lab measurements. * Performance data for this new function is planned to be published after general availability. * When you are using the new zSeries Application Assist Processor (zAAP), which provides a specialized z/OS Java™ execution environment, the total number of processors defined in a z/OS logical partition is the sum of general purpose processors (CPs) and zSeries zAAPs. -- John Eells z/OS Technical Marketing IBM Poughkeepsie [EMAIL PROTECTED] -- 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: z/OS 1.6 and more than 16 processors
... Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? ... Maybe yes and maybe no. So far nobody has bellied up to the bar with an answer, that I know of. But, unless the vendor supplies applications written in JAVA, I would think not. -teD In God we Trust! All others bring data! --Deming -- 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: z/OS 1.6 and more than 16 processors
Mark, Care to bet who will be the first? Ed Ed, I'm sure there are others, probably Mark as seem to recall gleaning some info from him during planning, but we implemented one in Feb. In fact, my PHB (just kidding Bruce) had me review his upcoming CMG presentation Getting zAAPed this week. Paul -- 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: z/OS 1.6 and more than 16 processors
... NO! At least not yet. That is the whole point. Well, almost the whole point. A zAAP engine is (IIRC)also about 1/4 the price of a general purpose processor. ... Also, they run at full speed. This only matters on z/890's with variable speeds. But, where they get you is on maintenance charges, once they kick in. At least in Canada. -teD In God we Trust! All others bring data! --Deming -- 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: z/OS 1.6 and more than 16 processors
On Thu, 7 Jul 2005 14:15:52 -0500, McKown, John wrote: -Original Message- On Behalf Of Tom Schmidt... snip I'm waiting for the first license agreement that simply demands a seat on the company's board of directors and stock options for the vendor. We can't be all that far away from that, can we? Well, if they are on the board, won't they be SOX candidates? Not likely that a vendor would want that. Stock options may not be all that attractive, depending on the licensee's stock price and history. SOX is a US-only issue and may not even be that after last week's CEO acquittal. Certain software companies have been accused of manipulating their own stock for fun profit... imagine if they could extend that value- added service through their customer base. -- Tom Schmidt Madison, WI -- 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: z/OS 1.6 and more than 16 processors
R.S. wrote: IMHO it is also worth to say about LPAR weights with zAAP aka IFA. I was told (it's not my observation) about some gotcha's when different types of processors (CP, IFC, IFL, IFA) are in use, especially when CF LPARs are in use (due to 'active wait'). Work is dispatched to zAAP processors by the MVS dispatcher in a manner similar to what's done for general purpose CPs. There is no active wait (a euphemism for hard CPU loop). -- - | Edward E. Jaffe|| | Mgr, Research Development| [EMAIL PROTECTED]| | Phoenix Software International | Tel: (310) 338-0400 x318 | | 5200 W Century Blvd, Suite 800 | Fax: (310) 338-0801| | Los Angeles, CA 90045 | http://www.phoenixsoftware.com | - -- 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: z/OS 1.6 and more than 16 processors
Edward E. Jaffe wrote: R.S. wrote: IMHO it is also worth to say about LPAR weights with zAAP aka IFA. I was told (it's not my observation) about some gotcha's when different types of processors (CP, IFC, IFL, IFA) are in use, especially when CF LPARs are in use (due to 'active wait'). Work is dispatched to zAAP processors by the MVS dispatcher in a manner similar to what's done for general purpose CPs. There is no active wait (a euphemism for hard CPU loop). Yes, but, as Mark pointed, they are in common weight pool with IFC's. -- 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
Re: z/OS 1.6 and more than 16 processors
On Jul 7, 2005, at 2:13 PM, Paul Dineen wrote: Just curious... do the vendors who charge fees by capacity take the zAAP processor into consideration? Ed Ed, IBM does recommend checking with your vendor, but there are no software charges as zAAP engine is JAVA specific. Now if you're vendor software is JAVA based, I'd guess you may need to pay for zAAP capacity. The hardware cost is $125/K per zAAP engine. Paul Paul, I was thinking more of an ISV charging for it. Its still pretty new (at least to me) so my thoughts are they (the ISVs) will be lurking around and maybe testing the waters. I am sure that the first to charge for it will raise a stink in the community. But they may just use a small no name company to see what they can get away with. I have a guess on which ISV will do it first, but who knows the company with deep pockets running Oricle might be the test case. Ed -- 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