1 A note about CPU utilization... 2 3 For HP-UX 11.0 <= system < 11.23 the configure script will select the 4 "pstat" CPU utilization mechanism. This mechanism is the familiar 5 HP-UX idle counter mechanism (for all incense and porpoises) and 6 requires calibration. See src/netcpu_pstat.c for all the details. 7 8 For HP-UX 11.23 >= system, the configure script will select the 9 "pstatnew" CPU utilization mechanism. 11.23 adds cycle counts for 10 user, kernel and interrupt modes to the idle cycle counter. As such, 11 it _should_ be possible to simply take the sum of the four and the 12 fractions and know how much time was spent in each mode. 13 HOWEVER... there is a bug in the accounting for interrupt cycles, 14 where interrupt cycles go missing. SOOO, since there is an accurate 15 way to know what the total number of cycles should have been over the 16 interval, and we know (ass-u-me) that the idle cycle counter is good 17 (since the pstat mechanism has tested that one OK), we will take the 18 ratio of idle to total cycles to compute CPU util. 19 20 We will still calculate fractions for user, kernel and interrupt, and 21 report them in debug (-d) output, but with a warning for interrupt 22 time. See src/netcpu_pstatnew.c for all the details. 23 24 Up through HP-UX 11.23 (aka 11iV2) if you enable burst mode, and 25 happen to send sub-MSS requests and/or responses you _cannot_ assume 26 that the packet per second rate on the wire will match the transaction 27 per second rate reported by netperf, even if you set TCP_NODELAY with 28 the test-specific -D option. The HP-UX 11.X TCP stack likely will be 29 generating some immediate 'standalone' ACKnowledgements which may not 30 be generated by other stacks. This has been reported to the HP-UX TCP 31 folks, and an announcement will be made when that issue is resolved. 32
