Re: 250+
"mfnoel" <mike-a/sRUk7aFfFWk0Htik3J/[email protected]>
| Newsgroups | gmane.comp.emulators.turnkey-mvs |
|---|---|
| Message-ID | <[email protected]> |
All running with the same priority. Just perform=2 as defaulted by JES2. You can see in the MF1 report that all the transactions were period 1. IPS/OPT unmodified. There was no attempt to "stress" the system per se; my goal was just to get 250+ running at all!! The ramp up was average 1 logon in 5 seconds (outside the window of the MF1 report). Transaction "think time" was average 30 seconds, so the transaction rate was a bit under 10 per second. Judging by the apparent extra capacity even on my old laptop I think that transaction rate could be increased significantly, but I'm not into "tuning" this configuration, just demonstrating it. In retrospect the most "stressful" aspect of the scripts is that they all hit the same few vsam files, so obviously contention for file access (enqueue waits) was significant. Some of the long worst case response times were (probably) due to the way KICKS does these enqueues (random wait, try again instead of just getting in line). --- In [email protected], "cedarlakebuzzards" <cedarlakebuzzards@...> wrote: > Each KICS session is really a TSO session, does each TSO session run > with the same dispatching priority? Were any of these 3270 scripts > designed to stress the system?