HP UX java program use 100% cpu

[zznode]

i have 5+ java program on my hp 7420.it have 8 processors 16 cores cpu.and now i find out one java program use 100% cpu since i replace some jar files.this is glance info:
B3692A Glance C.05.00.000 15:31:46 tnmsapp2 9000/800 Current Avg High
------------------------------------------------------------------------------------------------------------------------------------
Cpu Util S SU U | 98% 99% 100%
Disk Util F F | 4% 4% 5%
Mem Util S SU UB B | 61% 61% 61%
Networkil U UR R | 39% 39% 39%
------------------------------------------------------------------------------------------------------------------------------------
PROCESS LIST Users= 4
User CPU % Thrd Disk Memory Block
Process Name PID Name (1600% max) Cnt IOrate RSS/VSS On
--------------------------------------------------------------------------------
java 11275 tnms2 1298.0 625 16.2 4.28gb 9.01gb SLEEP
java 11280 tnms2 138.4 155 1.3 3.29gb 5.63gb STRMS
java 11293 tnms2 101.4 561 o/f 7.65gb 8.06gb STRMS
vxfsd 63 root 11.5 81 2.1 21.8mb 21.8mb SLEEP
midaemon 22295 root 4.3 2 0.0 66.1mb 84.8mb SLEEP
java 11270 tnms2 4.3 281 136.8 1.20gb 4.54gb SLEEP
java 11285 tnms2 0.4 176 0.1 1.17gb 1.71gb SLEEP
glance 22291 root 0.4 1 0.0 21.9mb 28.1mb STRMS
sh 22296 patrol 0.2 1 0.3 20.8mb 28.7mb died
java 11258 tnms2 0.2 122 0.1 955.9mb 1.23gb SLEEP
java 11263 tnms2 0.0 41 0.1 1.00gb 1.19gb SLEEP
PatrolAgent 9773 patrol 0.0 1 3.5 19.4mb 34.4mb OTHER
cmcld 16679 root 0.0 6 0.0 46.1mb 46.1mb SLEEP



pid 11275 result:
CPU Usage (util): 1382.2 Log Reads : na Wait Reason : PRI
User/Nice/RT CPU: 1343.9 Log Writes: na Total RSS/VSS : 4.23gb/ 9.01gb
System CPU : 11.7 Phy Reads : 0 Traps / Vfaults: 21079/ 24
Interrupt CPU : 22.5 Phy Writes: 71 Faults Mem/Disk: 0/ 0
Cont Switch CPU : 3.6 FS Reads : 0 Deactivations : 0
Scheduler : HPUX FS Writes : 71 Forks & Vforks : 0
Priority : 241 VM Reads : na Signals Recd : 0
Nice Value : 20 VM Writes : na Mesg Sent/Recd : 6689/ 7134
Dispatches : 53504 Sys Reads : na Other Log Rd/Wt: na/ na
Forced CSwitch : 21541 Sys Writes: na Other Phy Rd/Wt: na/ na
VoluntaryCSwitch: 31466 Raw Reads : na Proc Start Time
Running CPU : 7 Raw Writes: na Wed Dec 22 06:00:01 2010
CPU Switches : 7 Bytes Xfer: 2.2mb :
Argv1: 10.209.39.17
Cmd : /opt/java1.5/bin/PA_RISC2.0W/java -DFm


how can i trace the problem using hpux tools?

[DeafFrog]

Hi ,

what do those jar do , those you moved ?
1)check if you are up to date on patches :
hpjconfig , as per you server model.
http://docs.hp.com/en/5992-4687/ch01s06.html.
2)kill -3 to the java process , do a heap and dump anlysis .you can download the heasp and dump analyser frorm IBM site.this will let you know if you have waiting threads and the reason for the same.

Reg,

[zznode]

hi,DeafFrog

thanks very much.

i will try it.


best regards.

[Emil Velez]

download hpjmeter (do a search at software.hp.com) for it. Install it and start using to to analyse your java programs