I understand that prstat -m shows microstate wallclock utilization that should 
add up to 100% from USR to LAT. However, I get several entries that clearly do 
not add up. See below for PID 18022 as an example.

   PID USERNAME USR SYS TRP TFL DFL LCK SLP LAT VCX ICX SCL SIG PROCESS/NLWP  
  5859 root     1.3 4.5  -   -   -   -   93  -    3 128  4K   0 esd/1
 24295 cemsbin  2.0 2.5  -   -   -   -  0.0  -   50  50 23K   0 alarmifproce/1
 15464 root     2.5 1.0  -   -   -   -   32  -   27  16 634  17 ksh/1
     1 root     1.0 2.0  -   -   -   -  0.0  -   66  11  1K  71 init/1
 13183 cemsadm  0.2 2.0 0.0 0.0 0.0 0.0  98 0.0  31   4 961   1 prstat/1
  1357 root     0.7 0.0  -   -   -   -  0.0  -   73   4 108   0 esd/2
 19994 root     0.0 0.5  -   -   -   -  0.0  -    8   0   0   0 pwait/1
 10405 root     0.3 0.0  -   -   -   -   28  -  427  36  3K   0 nco_objserv/36
 11925 root     0.1 0.0  -   -   -   -   63  -  113  56 962   0 nco_p_socket/110
 18022 cemsbin  0.0 0.0  -   -   -   -   59  -    1   0   4   0 java/141
  5836 root     0.0 0.0  -   -   -   -   81  -    2   0   5   0 picld/21
 11978 cemsbin  0.0 0.0  -   -   -   -  100  -    2   0   9   0 oracle/1

When I use ptstat -mL, I get similar results (see the second line):
   PID USERNAME USR SYS TRP TFL DFL LCK SLP LAT VCX ICX SCL SIG PROCESS/LWPID 
 15155 nortel    95 4.2  -   -   -   -  0.6  -    8 672 12K   0 compress/1
  8330 nortel   2.6 9.6  -   -   -   -  0.0  -   1K  44  9K   0 wam_zapOneGr/1
     1 root     0.8 2.3  -   -   -   -  0.0  -  484  14 14K 692 init/1
  1357 root     2.7 0.2  -   -   -   -  0.0  -  214  20  1K   0 esd/1
  8315 nortel   0.4 2.2  -   -   -   -  0.0  -  293  14  2K   0 wam_zapOneGr/1
 14093 cemsadm  0.5 1.6 0.1 0.0 0.0 0.0  98 0.0  23  12  2K   0 prstat/1
 12881 root     0.2 0.2  -   -   -   -  0.0  -   10   3  57   0 java/102

Why?
 
 
This message posted from opensolaris.org
_______________________________________________
perf-discuss mailing list
perf-discuss@opensolaris.org

Reply via email to