Thanks! As long as I can trust (more or less) the numbers for the USR/SYS and avoid the dampening effect of the regular prstat - I am happy.
I run a little test comparing "prstat -m 1" to "vmstat" 1. I took about 300 samples each starting at about the same time. Then I added USR and SYS percentages for all processes for each sample and divided the result by 4 (number of CPUs). I plotted this against vmstat cpu_us - and the result was not identical but very close. Essentially: PRSTAT:SUMofALL(SYS+USR)/Nprocessors = VMSTAT:cpu_us Is this what one would expect? -eM -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, February 12, 2007 5:32 PM To: Margulis, Eugene (CAR:2Y82) Cc: perf-discuss@opensolaris.org Subject: Re: [perf-discuss] Re: Re: prstat-m - cpu% does not add to 100% ? > Perfect - this answers my question! No, I don't want to imply that the > mechanism is broken - as long as I understand how to read the data and > how to interpret the discrepancies. With this problem, the data is difficult to interpret on Solaris 9 machines. If you were to request a bug to be filed on this, I would suggest it be against against procfs, not microstate accounting. However, if you want to actually be able to obtain reasonable numbers immediately, the problem has already been fixed on Solaris 10. I'm running ONNV and my output looks nice like this: PID USERNAME USR SYS TRP TFL DFL LCK SLP LAT VCX ICX SCL SIG PROCESS/NLWP 3067 johansen 1.0 0.2 0.0 0.0 0.0 33 65 0.1 1K 123 10K 0 firefox-bin/3 1723 johansen 0.6 0.3 0.0 0.0 0.0 0.0 99 0.2 1K 10 5K 141 Xorg/1 3301 johansen 0.0 0.1 0.0 0.0 0.0 0.0 100 0.0 45 0 404 0 prstat/1 2474 johansen 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 51 0 461 51 xlp/1 3298 johansen 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 3 1 57 0 xterm/1 2480 johansen 0.0 0.0 0.0 0.0 0.0 0.0 100 0.1 148 0 295 0 asclock/1 275 root 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 2 0 53 0 utmpd/1 217 daemon 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 1 0 8 0 rpcbind/1 2487 johansen 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 14 1 10 0 tail/1 3024 johansen 0.0 0.0 0.0 0.0 0.0 86 14 0.0 209 1 173 0 java/14 531 root 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 3 0 10 0 sendmail/1 530 smmsp 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 0 0 0 0 sendmail/1 554 root 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 0 0 0 0 fbconsole/1 2445 johansen 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 0 0 0 0 bash/1 491 root 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 0 0 0 0 sshd/1 3070 johansen 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 0 0 0 0 sh/1 425 root 0.0 0.0 0.0 0.0 0.0 20 80 0.0 0 0 0 0 automountd/5 452 root 0.0 0.0 0.0 0.0 0.0 62 38 0.0 0 0 0 0 syslogd/13 Notice that this isn't exactly perfect, (firefox only totals 99.3 and Xorg is 100.1) but it is much closer to 100 %. The LAT numbers are also possibly inflated due to CR 6498304, which was fixed in ONNV build 57. -j _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
