We're not hitting the storage issues Jim and Rick are. Our 4 6-engine LPARs are happily on 32G each with zero-little paging... long way to go as far as running out of storage. And that may be because we have some CPU intensive workload skewing us the other way. We also squeeze and tune like crazy our test/dev environments because they are traditionally short changed on storage ... I think at one point we were 5-6:1 on dev/test (100 servers running on 24 gig). Better now, but still probably 2.5:1 at the moment (economies of scale play out.. Many devs groups can't all be active at the same time).
It is the administrative overhead of z/VM management that I am worried about. I have to justify that against whatever capacity is lost due to more systems/lpars. Storage concerns aren't close to a concern yet. Marcy "This message may contain confidential and/or privileged information. If you are not the addressee or authorized to receive this for the addressee, you must not use, copy, disclose, or take any action based on this message or any information herein. If you have received this message in error, please advise the sender immediately by reply e-mail and delete this message. Thank you for your cooperation." -----Original Message----- From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf Of Mark Post Sent: Saturday, August 08, 2009 5:10 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: [IBMVM] MP effect on z/VM Linux hosting >>> On 8/8/2009 at 12:07 PM, Marcy Cortes <marcy.d.cor...@wellsfargo.com> >>> wrote: -snip- > The real concern is whether to put 6 IFLS on 7 different boxes or do we lose > anything by having say 14 IFLs on 3 boxes or 10-10-11-11 on 4 boxes. Based on what Jim Vincent and Rick Barlow have been saying for some time, and adding in what Barton Robinson just wrote, I would say that you're likely to run out of real storage before you run into too much MP effect. In your case, I would say you would need to balance the administrative effort of managing X z/VM systems on Y LPARs against how much real storage you need and how many CPUs you need. I would personally tend to go with fewer "larger" z/VM instances, but that may not be a concern for you. Mark Post