Any LPAR can exceed their proportion if there is available resource. If there is insufficient resource, LPARS will be forced to their proportion, as defined by the LPAR weights (unless another LPAR is not consuming "its share").
Remember, PR/SM does not know LPARA is "production" and should be favored. It just knows LPARA is assigned weight x. For example there is CEC with LPARS(weights) LPARA(60%) LPARB(20%) and LPARC(20%). If all LPARS demand their weight, the CPU will be distributed according to the weights defined (barring soft/hard capping). Now, assume LPARC only wants 5%. LPARA and LPARB can now consume the 15% that LPARC does not want. In the example from my previous post, the total demand is 105%. LPARB has a defined weight of 20% and is guaranteed this value. LPARA is guaranteed its weight of 80%. (total of 100). The additional 5% exceeds the available capacity and will be presented as "latent demand". IF LPARB were to drop to only requiring 10%, LPAR would then be allowed to exceed the defined proportion because the total demand is now 90%. There are man7y "tricks of the trade" to "cap" the non-production LPARs. The only question is the granularity appropriate. Another good source would be the CMG archives (www.cmg.org) and Cheryl Watsons Tuning Letters. Hopefully, Cheryl will spot this thread and chime in. HTH, <snip> Can you expand on this example: If LPARA wants 85% and LPARB want 20% (total 105%) LPARB will get 20% and LPARA will be squeezed to 80%. It seems counter intuitive to me and I'd like to understand. Lets say LPARA is prod - they should get most of the resources. Why would LPARA be squeezed instead of LPARB? </snip> ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
