On Wed, 12 Sep 2007 19:05:19 +0000, Ted MacNEIL wrote: >>I see no reason why PR/SM can't support more logical processors for a partition than there are physical processors > >It was a design decision, from the get go. >MDF was designed the same way. > >I was a beta-tester of MDF in the mid-1980's, and participated >in many NDA meetings for that.
Ok, and I was a cross-trained Amdahl SE Specialist at the time. I was deeply involved it the 580, macrocode and MDF > >Also, attended many NDA's for PR/SM. >The company, where I worked at the time, was an early adopter of PRSM. > >Both IBM and Amdahl were quick to point out the design decision, >and stated that if you wanted to go the other way (more >logical/image than physically install), use VM. True. That doesn't mean it can't be done. You can't define a partition with more active logical processors than there are on the box. Neither can you configure another one online when you already have as many as there are physical processors. That begs the question, though. How do you explain the facts as reported by Mr. Merritt? My suspicion is that PR/SM doesn't really care that much that he is left with four logical processors while funning on a box with only three CPs. I don't think I'd want to keep it running that way, but what else is PR/SM going to do? Stop dispatching on of the logical processors? And if so, what happens to the task that was running on it at the time? -- Tom Marchant ---------------------------------------------------------------------- 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
