CPU usage is very high due to the patrol and STM

[HP UNIX Professionals]

Hi Experts,

We have a HPUNIX box(B.11.00 U 9000/800), Patrol Agent installed for Monitoring.

Due to this patrol Agent we have server performance problem .

This agent is stimulating pvdisplay, vdisplay such as command through STM.

Please guide me to sort out this issue.


Top
===
root@xxxx:/# top
System: xxxxWed Jan 5 14:35:32 2011
Load averages: 10.24, 9.90, 8.71
404 processes: 308 sleeping, 37 running, 57 zombies
Cpu states:
CPU LOAD USER NICE SYS IDLE BLOCK SWAIT INTR SSYS
0 10.33 16.8% 0.0% 83.2% 0.0% 0.0% 0.0% 0.0% 0.0%
1 9.72 17.8% 0.0% 82.2% 0.0% 0.0% 0.0% 0.0% 0.0%
2 10.68 14.9% 0.0% 85.1% 0.0% 0.0% 0.0% 0.0% 0.0%
3 10.11 11.9% 0.0% 88.1% 0.0% 0.0% 0.0% 0.0% 0.0%
4 10.20 17.8% 0.0% 82.2% 0.0% 0.0% 0.0% 0.0% 0.0%
5 10.37 32.7% 0.0% 67.3% 0.0% 0.0% 0.0% 0.0% 0.0%
--- ---- ----- ----- ----- ----- ----- ----- ----- -----
avg 10.24 18.8% 0.0% 81.2% 0.0% 0.0% 0.0% 0.0% 0.0%
Memory: 638316K (534384K) real, 676492K (596164K) virtual, 3771572K free Page# 1/14
CPU TTYPID USERNAME PRI NI SIZE RES STATE TIME %WCPU %CPU COMMAND
5 pts/td 25471 etrader 229 20 3408K 1872K run0:02 4.38 4.36 egrep
5 ? 11089 oracle 229 20 1564M 1988K run0:10 4.30 4.30 oracleBKHW
3 ?2326 root212 20 4860K 496K run 17132:53 3.25 3.25 uvdlockd
2 ? 24367 root203 20 2576K 996K run0:01 2.17 2.16 pvdisplay
1 ? 22564 root195 20 2576K 996K run0:01 1.51 1.51 pvdisplay
2 ? 24709 root203 20 2576K 996K run0:01 1.46 1.46 pvdisplay
3 ?2085 root185 20 7056K0K run0:00 13.34 1.27 vgdisplay
2 ?38 root152 20 1856K0K run46:37 1.27 1.27 vxfsd
1 ?2040 root184 20 6864K0K run0:00 12.31 1.17 vgdisplay
1 ?2005 root184 20 7056K 864K run0:00 8.01 1.12 vgdisplay
2 ?2107 root185 20 6864K0K run0:00 21.02 1.03 vgdisplay
2 ?683 root154 20 1676K 160K sleep 253:39 0.93 0.93 syncer
0 ?2080 root183 20 6416K0K run0:00 8.58 0.82 vgdisplay
1 ?2055 root184 20 6864K0K run0:00 7.70 0.73 vgdisplay
2 ?2745 root179 20 32K0K zomb0:00 0.68 0.68 agdbserver
0 ?2087 root183 20 6672K0K run0:00 6.16 0.59 vgdisplay
0 ?2117 root183 20 6224K0K run0:00 12.01 0.59 vgdisplay
0 ?2106 root181 20 5264K0K run0:00 11.01 0.54 vgdisplay
0 ?2121 root183 20 6864K0K run0:00 11.01 0.54 vgdisplay
0 ?2119 root181 20 5840K0K run0:00 10.01 0.49 vgdisplay



root@xxxxx:/# ps -ef | grep patrol
patrol 25648 1 52 Dec 14 ? 548:35 PatrolAgent -p 3181
patrol 26145 25648 0 11:24:50 ? 0:00 /usr/bin/ksh
patrol 7557 25648 0 08:45:13 ? 0:00 /usr/bin/sh /bin/stm -c
patrol 7560 7559 0 08:45:14 ? 1:08 stm_map_read 0 dummy /var/tmp/stm7559/mapread.cfg
root 20334 12595 0 15:15:24 pts/ta 0:00 grep patrol
patrol 20332 25648 0 15:15:20 ? 0:00 vmstat 5 2
patrol 7679 7559 0 08:45:58 ? 0:00 sh -c cat /var/tmp/stm7559/tmp/baaa07559

Thanks in advance

vijnay

[Mel Burslan]

Short of knowing why the patrol agent spawns that many lvm related commands, you can not do much about it but inquire about to the patrol vendor to see if there is a more up-to-date agnet for your platform. But considering you are at v.11.0, I highly doubt that. I am pretty sure even that vendor must have deprecated the support for their products running on hpux v.11.0
Consider upgrading...

[Dennis Handly]

Who is the parent of all of the *display commands:
UNIX95=EXTENDED_PS ps -H -fu root

[HP UNIX Professionals]

Hi Deny,

Please tell me how to find the parent of â displayâ commands

Todays top output
=================
ystem: xxxx Wed Jan 5 17:00:23 2011
Load averages: 7.54, 4.86, 4.11
575 processes: 411 sleeping, 82 running, 82 zombies
Cpu states:
CPU LOAD USER NICE SYS IDLE BLOCK SWAIT INTR SSYS
0 8.13 18.0% 0.0% 82.0% 0.0% 0.0% 0.0% 0.0% 0.0%
1 7.57 21.3% 0.0% 78.7% 0.0% 0.0% 0.0% 0.0% 0.0%
2 7.12 20.3% 0.2% 79.5% 0.0% 0.0% 0.0% 0.0% 0.0%
3 7.59 24.6% 0.2% 75.2% 0.0% 0.0% 0.0% 0.0% 0.0%
4 7.85 19.3% 0.0% 80.7% 0.0% 0.0% 0.0% 0.0% 0.0%
5 7.01 20.7% 0.0% 79.3% 0.0% 0.0% 0.0% 0.0% 0.0%
--- ---- ----- ----- ----- ----- ----- ----- ----- -----
avg 7.54 20.7% 0.2% 79.1% 0.0% 0.0% 0.0% 0.0% 0.0%

Memory: 922620K (803152K) real, 992272K (896408K) virtual, 3491900K free Page# 1/14

CPU TTY PID USERNAME PRI NI SIZE RES STATE TIME %WCPU %CPU COMMAND
5 ? 1789 root 154 10 4584K 1364K sleep 11:23 5.20 5.19 diagmond
0 ? 29924 patrol 186 20 11800K 1632K run 0:02 1.76 1.76 stm
5 ? 2326 root 211 20 4860K 496K run 17257:49 1.66 1.65 uvdlockd
5 ? 38 root 152 20 1856K 0K run 46:50 1.41 1.40 vxfsd
0 ? 9177 root 184 20 7312K 0K run 0:00 10.05 1.40 vgdisplay
4 ? 9143 root 185 20 7312K 0K run 0:00 9.77 1.36 vgdisplay
0 ? 9191 root 185 20 7312K 0K run 0:00 9.44 1.31 vgdisplay
3 ? 9047 root 185 20 7056K 544K run 0:00 7.17 1.30 vgdisplay
4 ? 9141 root 183 20 7056K 0K run 0:00 9.27 1.29 vgdisplay
5 ? 9186 root 184 20 7056K 0K run 0:00 9.20 1.28 vgdisplay
0 ? 9187 root 185 20 6864K 0K run 0:00 9.09 1.27 vgdisplay
0 ? 9137 root 184 20 7056K 0K run 0:00 9.04 1.26 vgdisplay
4 ? 9207 root 185 20 6864K 0K run 0:00 9.00 1.25 vgdisplay
2 ? 8970 root 184 20 7312K 5604K run 0:00 4.78 1.24 vgdisplay
1 ? 9093 root 184 20 6864K 0K run 0:00 6.81 1.23 vgdisplay
0 ? 9037 root 184 20 7056K 5412K run 0:00 5.57 1.23 vgdisplay
1 ? 9129 root 184 20 7056K 0K run 0:00 8.60 1.20 vgdisplay
1 ? 9147 root 184 20 7056K 0K run 0:00 8.58 1.20 vgdisplay
2 ? 9015 root 184 20 7056K 4836K run 0:00 5.36 1.19 vgdisplay
3 ? 9119 root 184 20 6864K 0K run 0:00 8.48 1.18 vgdisplay
5 ? 9231 root 184 20 6672K 0K run 0:00 12.42 1.18 vgdisplay
3 ? 9055 root 184 20 6864K 3684K run 0:00 6.50 1.18 vgdisplay
1 ? 9027 root 185 20 6864K 5284K run 0:00 5.32 1.18 vgdisplay

[Mel Burslan]

Please tell me how to find the parent of â  displayâ  commands

Do you se the command at the end of Dennis' post ? This one below:

UNIX95=EXTENDED_PS ps -H -fu root

it will give you the information you need. PPID (parent process ID) is the third column in the output you are going to get from this command

[Alzhy]

UNIX95=1 ps -efH

It lists the entire process tree in genealogical or TREE fashion... should be easy to figure out

You probably have a very busy system or there is something wrong with your disks that LVM vgdisplay commands are piling up...