> -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of > Martin Packer > Sent: Wednesday, April 20, 2011 3:14 AM > To: [email protected] > Subject: [IBM-MAIN] Dyadic vs AP: Was "CPU utilization/forecasting" > > Ron, care to remind us of the modelling difference? It's been a while. :-) > [Ron Hawkins] Martin,
I'm calling in a lot of swapped out grey matter here, but if I recall correctly with the AP you would not calculate a probability of a CPU being busy as a function of the two processors. The scheduling in AP was treated separately from the main CPU. I remember spending some time playing with CPU affinity trying to keep the CPU bound jobs away from the AP and that DUO (UCC2) required a CPUAFF=1 which made the AP a pain in the butt after converting from DOS. Someone correct me please, but didn't it require a swap to move from the AP to Main Processor to issue an IO. So Martin, not matter how relatively incorrect my recollections are, the 3033 AP required an entirely different approach to tuning and capacity planning than a 390-200, where CPU queuing was usually a simple function of PCTCPBY0 x PCTCPBY1. Ron ---------------------------------------------------------------------- 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
