Hi Allan, I wish I had this information, but I don't. I agree with you that the "beans are merely counted differently." And I agree with Kenneth about "It is what it is." In other words, I would only be concerned if it dramatically changed one day.
Best regards, Cheryl ====================== Cheryl Watson Watson & Walker, Inc. www.watsonwalker.com cell & text: 941-266-6609 ====================== On May 14, 2014, at 11:20 AM, Staller, Allan wrote: *ix systems in general, spend a higher proportion of time in "supervisor state" than z/OS. This is not a bad thing. The beans are merely counted differently. Much of what z/OS and is predecessors manage to charge back to a specific user is not done in *ix. CMG probably has some good reference materiel in the archives. Cheryl Watson would also have some good info. ISTR some old numbers (without attribution) of +/- 50% supervisor state for *ix systems and +/- 10% for z/OS and predecessors.** **These numbers are about 10 years old may no longer be accurate. <snip> > On 5/14/2014 7:23 AM, Klein, Kenneth E wrote: >> What is the general opinion out there on what portion of CPU cycles should >> be spent in supervisor state? >> Is more than 50% a "bad thing"? > > I can't imagine >50% is either good or bad. It is what it is; entirely > dependent on the nature of the programs you are running at any given time. > It may also depend on the OS. I suspect Linux spends a smaller portion in supervisor state than z/OS. </snip> ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
