WebSphere caused the recommendation of the 2 CPU setup. I also appreciate all the educations on SET SHARE vs. SET CPU.
I am now contemplating which cliff to jump off, if not both. -----Original Message----- From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf Of RPN01 Sent: Thursday, July 08, 2010 11:24 AM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: Define CPU's Another way of wording this is that adding CPUs to a virtual machine allows the guest to take advantage of multitasking, but does not increase the total amount of CPU time the image receives. If single threading tasks in the image is the bottleneck, then adding a CPU may relieve it. But if this isn't the issue, then it won't help. Now, I think I saw WebSphere mentioned somewhere along the line, and I think that it will take advantage of multitasking, given the increased number of available CPUs. -- Robert P. Nix Mayo Foundation .~. RO-OC-1-18 200 First Street SW /V\ 507-284-0844 Rochester, MN 55905 /( )\ ----- ^^-^^ "In theory, theory and practice are the same, but in practice, theory and practice are different." On 7/8/10 10:20 AM, "Alan Altmark" <alan_altm...@us.ibm.com> wrote: > On Thursday, 07/08/2010 at 11:02 EDT, Martin Zimelis > <martin.zime...@gmail.com> wrote: >> Unless you're max'ing out these virtual machines by consuming 100% >> of a real processor, it should be as simple as increasing their SHARE >> values. > > To finish the thought, adding virtual CPUs to a guest does not add CPU > capacity to a guest; SET SHARE does. Adding another virtual CPU may allow > the guest to better use the CPU capacity it has been given, increasing > throughput or decreasing response time, or it may actually slow the guest > down. It all depends on the application. > > A good performance monitor will tell you if a guest is constrained, and > why. Of course, one must measure, change, and measure again to ensure > that the changes had the desired effect. Sometimes after you release the > hounds, you discover that the yard is a mess. > > Alan Altmark > z/VM Development > IBM Endicott ----------------------------------------------------- Please see the following link for the BlueCross BlueShield of Tennessee E-mail disclaimer: http://www.bcbst.com/email_disclaimer.shtm
