Re: Mixed page volume sizes
On Thu, Sep 30, 2010 at 12:53 AM, O'Brien, Dennis L dennis.l.o'br...@bankofamerica.com wrote: [ snip ] need them, so if I add another 32 GB of storage, my choice would be something like 212 3390-3's and 28 3390-9's, or 100 3390-9's. Which would be a better choice? As you indicated, z/VM spreads paging effectively to fill volumes with the same amount of data. This will fill your -3's quicker than the -9's. The risk is that at some point you filled the small ones completely and end up paging only with a subset of the subchannels. Mixing sizes is not recommended; depending on the actual numbers it may be a very bad idea (tm). Bottom line is to look at your performance monitor and determine whether users are unreasonably held back by paging and whether that troubles your users. Let's do the math: once you have filled 50% of your paging space (~ 340 GB) each will hold ~1.4 GB. So the -3's are at 60% and the -9's are at 20%. You'd still have 240 packs more or less contribute in paging. When you convert to all -9, you only have 100 packs doing the paging. So in this case, 240 is more than 100 ;-) This is assuming modern DASD, not the true 3390-9 that were rotating slower so you could stuff more bytes on a track (and take 3 times as long to read and write the data). If you plotted the dates when you added page packs, you may be able to predict when you post again... ;-) and knowing that you will run out of the maximum number of CP owned volumes real soon... suppose you would go for 28 extra 3390-27 instead. Once you fill that space for 50%, your -3's are stuffed and you only have space on the big ones, so you run with only ~10% of your paging volumes (that counts as bad idea). Would you have started with all -27's, you now had about twice the number of page packs working for you. PS I'm not making up these examples. We did see a customer convert to just a few very big page devices. They were not happy. Rob -- Rob van der Heij Velocity Software http://www.velocitysoftware.com/
Re: Mixed page volume sizes
Dennis, I would add one more thing to what Rob said. Your paging rate should also be taken into account. If 100 spindles will handle your peak paging rates without causing measurable page wait, I'd vote for going the all-3390-9 route rather than the -3 approach. And record one more vote that mixed sizes is a very bad idea (for the reasons Rob described). Marty On Thu, Sep 30, 2010 at 4:32 AM, Rob van der Heij rvdh...@velocitysoftware.com wrote: On Thu, Sep 30, 2010 at 12:53 AM, O'Brien, Dennis L dennis.l.o'br...@bankofamerica.com wrote: [ snip ] need them, so if I add another 32 GB of storage, my choice would be something like 212 3390-3's and 28 3390-9's, or 100 3390-9's. Which would be a better choice? As you indicated, z/VM spreads paging effectively to fill volumes with the same amount of data. This will fill your -3's quicker than the -9's. The risk is that at some point you filled the small ones completely and end up paging only with a subset of the subchannels. Mixing sizes is not recommended; depending on the actual numbers it may be a very bad idea (tm). Bottom line is to look at your performance monitor and determine whether users are unreasonably held back by paging and whether that troubles your users. Let's do the math: once you have filled 50% of your paging space (~ 340 GB) each will hold ~1.4 GB. So the -3's are at 60% and the -9's are at 20%. You'd still have 240 packs more or less contribute in paging. When you convert to all -9, you only have 100 packs doing the paging. So in this case, 240 is more than 100 ;-) This is assuming modern DASD, not the true 3390-9 that were rotating slower so you could stuff more bytes on a track (and take 3 times as long to read and write the data). If you plotted the dates when you added page packs, you may be able to predict when you post again... ;-) and knowing that you will run out of the maximum number of CP owned volumes real soon... suppose you would go for 28 extra 3390-27 instead. Once you fill that space for 50%, your -3's are stuffed and you only have space on the big ones, so you run with only ~10% of your paging volumes (that counts as bad idea). Would you have started with all -27's, you now had about twice the number of page packs working for you. PS I'm not making up these examples. We did see a customer convert to just a few very big page devices. They were not happy. Rob -- Rob van der Heij Velocity Software http://www.velocitysoftware.com/
Re: Mixed page volume sizes
On Thu, Sep 30, 2010 at 1:35 PM, Martin Zimelis martin.zime...@gmail.com wrote: And record one more vote that mixed sizes is a very bad idea (for the reasons Rob described). But your very bad may vary ;-) For anyone still awake, ponder on this one... Rob, we just put in 30 volumes 3390-9 for paging. After the IPL, we noticed that we failed to remove 540PAG (a 3390-3) from the CP owned list. Can't get it off because it is in use. Somebody said that mixing volume sizes is a very bad idea. Is it good enough to just drain that volume or should we plan for another IPL to take it off? | Rob
Re: Mixed page volume sizes
Dennis, You should be fine mixing sizes so long as you watch utilizations of individual volumes (particularly the smallest ones) and don't let any of them go over 50% full. We ran out of 3390-9 volumes a while ago and had to start adding 3390-27's for paging. Since they're all virtualized inside the storage array, device response times are pretty much all the same. When the 3390-9's get toward 50% full (or the high-water mark hits the end of the device - another thing I like to watch), I add some more 3390-27's. I pretty much disregard the aggregate page volume utilization, which is often less than 20% in our case. Given that 240-248 puts you very close to the max number of page volumes, 3390-3's may not be the best choice because with a bit more growth, the only thing you'll be able to do is watch space utilization go beyond 50%, which won't be good. A place to start may be 120 3390-9's @ 33% util each. Then watch individual device response times to see if you need more actuators. Good luck! Mark Wheeler UnitedHealth Group -- Excellence. Always. If Not Excellence, What? If Not Excellence Now, When? Tom Peters, author of The Little BIG Things Date: Wed, 29 Sep 2010 15:53:25 -0700 From: dennis.l.o'br...@bankofamerica.com Subject: Mixed page volume sizes To: IBMVM@LISTSERV.UARK.EDU On of our z/VM 5.4.0 systems is about to grow to 140 GB of storage. Given our target overcommit ratio of 3:1, and IBM's advice to keep paging space no more than 50% full, this should just about fit onto 240 or 248 3390-3 sized volumes. My question is what will happen the next time we add storage. Clearly, we'll have to start using 3390-9 sized volumes for paging. Would we be better off converting only enough volumes to satisfy the space requirement, which would maximize the number of devices, or should we convert all volumes to 3390-9 to keep them the same size? My concern is that if we mix sizes, CP will try to allocate the same amount of space on each volume, and the 3390-3's will get more than half full. On the other hand, maximizing the number of devices maximizes the number of concurrent I/O's. Our Storage people aren't going to give me 240 3390-9's if I don't need them, so if I add another 32 GB of storage, my choice would be something like 212 3390-3's and 28 3390-9's, or 100 3390-9's. Which would be a better choice? Dennis Decision is not a verb.
Re: Mixed page volume sizes
Sorry for my ignorance and naivete, but I am simply staggered at the volume of page volumes of whatever size, 100 - 240. I have just a Dev Test z/VM environment with 2 measly 3390-3's for Paging. Granted, I am running only 5 z/OS vm's and 3 Linux vm's. But, what does someone have to run that needs 240 page volumes? I am sure it is justified, I would just like to know more about the workload. Mark Wheeler mwheele...@hotmail.com Sent by: The IBM z/VM Operating System IBMVM@LISTSERV.UARK.EDU 09/30/2010 09:00 AM Please respond to The IBM z/VM Operating System IBMVM@LISTSERV.UARK.EDU To IBMVM@LISTSERV.UARK.EDU cc Subject Re: Mixed page volume sizes Dennis, You should be fine mixing sizes so long as you watch utilizations of individual volumes (particularly the smallest ones) and don't let any of them go over 50% full. We ran out of 3390-9 volumes a while ago and had to start adding 3390-27's for paging. Since they're all virtualized inside the storage array, device response times are pretty much all the same. When the 3390-9's get toward 50% full (or the high-water mark hits the end of the device - another thing I like to watch), I add some more 3390-27's. I pretty much disregard the aggregate page volume utilization, which is often less than 20% in our case. Given that 240-248 puts you very close to the max number of page volumes, 3390-3's may not be the best choice because with a bit more growth, the only thing you'll be able to do is watch space utilization go beyond 50%, which won't be good. A place to start may be 120 3390-9's @ 33% util each. Then watch individual device response times to see if you need more actuators. Good luck! Mark Wheeler UnitedHealth Group -- Excellence. Always. If Not Excellence, What? If Not Excellence Now, When? Tom Peters, author of The Little BIG Things Date: Wed, 29 Sep 2010 15:53:25 -0700 From: dennis.l.o'br...@bankofamerica.com Subject: Mixed page volume sizes To: IBMVM@LISTSERV.UARK.EDU On of our z/VM 5.4.0 systems is about to grow to 140 GB of storage. Given our target overcommit ratio of 3:1, and IBM's advice to keep paging space no more than 50% full, this should just about fit onto 240 or 248 3390-3 sized volumes. My question is what will happen the next time we add storage. Clearly, we'll have to start using 3390-9 sized volumes for paging. Would we be better off converting only enough volumes to satisfy the space requirement, which would maximize the number of devices, or should we convert all volumes to 3390-9 to keep them the same size? My concern is that if we mix sizes, CP will try to allocate the same amount of space on each volume, and the 3390-3's will get more than half full. On the other hand, maximizing the number of devices maximizes the number of concurrent I/O's. Our Storage people aren't going to give me 240 3390-9's if I don't need them, so if I add another 32 GB of storage, my choice would be something like 212 3390-3's and 28 3390-9's, or 100 3390-9's. Which would be a better choice? Dennis Decision is not a verb.
Re: Mixed page volume sizes
Toto, I've a feeling we're not in Kansas any more Rob van der Heij rvdh...@gmail.com Sent by: The IBM z/VM Operating System IBMVM@LISTSERV.UARK.EDU 09/30/2010 10:24 AM Please respond to The IBM z/VM Operating System IBMVM@LISTSERV.UARK.EDU To IBMVM@LISTSERV.UARK.EDU cc Subject Re: Mixed page volume sizes On Thu, Sep 30, 2010 at 3:39 PM, George Henke/NYLIC george_he...@newyorklife.com wrote: Sorry for my ignorance and naivete, but I am simply staggered at the volume of page volumes of whatever size, 100 - 240. I have just a Dev Test z/VM environment with 2 measly 3390-3's for Paging. Granted, I am running only 5 z/OS vm's and 3 Linux vm's. My test system does not have that either. But think of 250G real memory and some 100 Linux guests of 4-10 GB each, all running enterprise applications. | Rob
Re: Mixed page volume sizes
S if you had guests averaging 18GB each, and you follow recommendations for page volume utilization (50% as I understand it), Mod 9s would yield around 4GB useable page space each. That would give you 100 such images per VM .. (250 volumes times the 4GB/volume divided by 18GB per image)? I know this rough but am I headed in the right direction? I'm making the assumption (an this may be incorrect) that the available real storage backing could support virtual requirements. On 9/30/10 9:24 AM, Rob van der Heij rvdh...@gmail.com wrote: On Thu, Sep 30, 2010 at 3:39 PM, George Henke/NYLIC george_he...@newyorklife.com wrote: Sorry for my ignorance and naivete, but I am simply staggered at the volume of page volumes of whatever size, 100 - 240. I have just a Dev Test z/VM environment with 2 measly 3390-3's for Paging. Granted, I am running only 5 z/OS vm's and 3 Linux vm's. My test system does not have that either. But think of 250G real memory and some 100 Linux guests of 4-10 GB each, all running enterprise applications. | Rob --. .- .-. -.-- Gary Dennis Mantissa Corporation 1121 Edenton Street Birmingham, Alabama 35242-9257 0 ... living between the zeros... 0 p: 205.968-3942 m: 205.218-3937 f: 205.968.3932 gary.den...@mantissa.com http://www.mantissa.com http://www.idovos.com
Re: Mixed page volume sizes
In our case, we're talking about going from 76 GB to 140 GB. We're trying to stay ahead of requirements, so suggestions such as look at your performance monitor don't really work. The workload that will use the additional capacity hasn't arrived, yet. I need to submit my disk requirements for paging to our Storage team, because it takes them a month or two to add the disk. I can't afford to wait until performance gets bad. Right now, this system has about 30 Linux guests, all running WAS. Most of them have 6 GB of virtual storage. This is a development system, so it's actually larger than our production systems. Developers like to have several levels of testing (UAT, SIT, etc) for each production server. One of the drivers for the upgrade is a proposal to add a 28 GB Linux guest to development, and one to each of two production systems. The wisdom of having a single guest that large remains to be seen, as does the actual requirement for that much storage. I know about SET REORDER and all that. If they really do need that much storage, they might end up with multiple smaller guests. Dennis Decision is not a verb. -Original Message- From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf Of Gary M. Dennis Sent: Thursday, September 30, 2010 08:07 To: IBMVM@LISTSERV.UARK.EDU Subject: Re: [IBMVM] Mixed page volume sizes S if you had guests averaging 18GB each, and you follow recommendations for page volume utilization (50% as I understand it), Mod 9s would yield around 4GB useable page space each. That would give you 100 such images per VM .. (250 volumes times the 4GB/volume divided by 18GB per image)? I know this rough but am I headed in the right direction? I'm making the assumption (an this may be incorrect) that the available real storage backing could support virtual requirements. On 9/30/10 9:24 AM, Rob van der Heij rvdh...@gmail.com wrote: On Thu, Sep 30, 2010 at 3:39 PM, George Henke/NYLIC george_he...@newyorklife.com wrote: Sorry for my ignorance and naivete, but I am simply staggered at the volume of page volumes of whatever size, 100 - 240. I have just a Dev Test z/VM environment with 2 measly 3390-3's for Paging. Granted, I am running only 5 z/OS vm's and 3 Linux vm's. My test system does not have that either. But think of 250G real memory and some 100 Linux guests of 4-10 GB each, all running enterprise applications. | Rob --. .- .-. -.-- Gary Dennis Mantissa Corporation 1121 Edenton Street Birmingham, Alabama 35242-9257 0 ... living between the zeros... 0 p: 205.968-3942 m: 205.218-3937 f: 205.968.3932 gary.den...@mantissa.com http://www.mantissa.com http://www.idovos.com
Re: Mixed page volume sizes
CP will page out to the disk with the best performance, so yes, the mdl3's may be filled too much when mixing sizes. 2010/9/30 O'Brien, Dennis L dennis.l.o'br...@bankofamerica.comdennis.l.o%27br...@bankofamerica.com On of our z/VM 5.4.0 systems is about to grow to 140 GB of storage. Given our target overcommit ratio of 3:1, and IBM's advice to keep paging space no more than 50% full, this should just about fit onto 240 or 248 3390-3 sized volumes. My question is what will happen the next time we add storage. Clearly, we'll have to start using 3390-9 sized volumes for paging. Would we be better off converting only enough volumes to satisfy the space requirement, which would maximize the number of devices, or should we convert all volumes to 3390-9 to keep them the same size? My concern is that if we mix sizes, CP will try to allocate the same amount of space on each volume, and the 3390-3's will get more than half full. On the other hand, maximizing the number of devices maximizes the number of concurrent I/O's. Our Storage people aren't going to give me 240 3390-9's if I don't need them, so if I add another 32 GB of storage, my choice would be something like 212 3390-3's and 28 3390-9's, or 100 3390-9's. Which would be a better choice? Dennis Decision is not a verb. -- Kris Buelens, IBM Belgium, VM customer support
