spinning jsvc process LP 4.1
We're non clustered running RHAS3 with LP 4.1.1.28 build 28. A day ago a single jsvc process got into a state of continually running using up resources. For lack of a better term, a day later I now have 6 of those "spinning" jsvc processes. From the user standpoint the portal is responsive and works fine. Here's my top snapshot:
10:12:34 up 12 days, 6:50, 2 users, load average: 11.53, 11.62, 11.35
794 processes: 781 sleeping, 13 running, 0 zombie, 0 stopped
CPU states: cpu user nice system irq softirq iowait idle
total 91.0% 0.0% 6.7% 0.2% 1.8% 0.0% 0.0%
cpu00 91.0% 0.0% 6.4% 1.1% 1.3% 0.0% 0.0%
cpu01 90.4% 0.0% 9.1% 0.0% 0.3% 0.0% 0.0%
cpu02 93.7% 0.0% 3.9% 0.0% 2.3% 0.0% 0.0%
cpu03 89.0% 0.0% 7.6% 0.0% 3.3% 0.0% 0.0%
Mem: 10015464k av, 9932952k used, 82512k free, 0k shrd, 286104k buff
3203728k actv, 4993776k in_d, 231172k in_c
Swap: 2040244k av, 0k used, 2040244k free 7562348k cached
PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND
9503 admin 25 0 1436M 1.4G 40420 R 19.1 14.6 643:24 0 jsvc
9584 admin 25 0 1436M 1.4G 40420 R 17.6 14.6 1163m 1 jsvc
9567 admin 25 0 1436M 1.4G 40420 R 15.3 14.6 1072m 3 jsvc
9591 admin 25 0 1436M 1.4G 40420 R 11.6 14.6 810:24 2 jsvc
9616 admin 25 0 1436M 1.4G 40420 R 10.9 14.6 64:05 2 jsvc
9589 admin 25 0 1436M 1.4G 40420 R 10.7 14.6 1134m 1 jsvc
26925 admin 15 0 1436M 1.4G 40420 S 0.7 14.6 0:04 3 jsvc
31718 admin 16 0 2728 2728 880 R 0.5 0.0 0:00 0 top
8041 admin 15 0 104M 104M 42084 S 0.4 1.0 10:27 0 ns-slapd
9565 admin 16 0 1436M 1.4G 40420 S 0.4 14.6 2:42 2 jsvc
30140 admin 15 0 1436M 1.4G 40420 S 0.4 14.6 0:21 0 jsvc
7990 admin 15 0 104M 104M 42084 S 0.3 1.0 28:27 1 ns-slapd
8631 admin 13 -2 12760 12M 10656 S < 0.3 0.1 14:50 2 csadmind
26208 admin 15 0 1436M 1.4G 40420 S 0.3 14.6 0:03 0 jsvc
26210 admin 15 0 1436M 1.4G 40420 S 0.3 14.6 0:02 3 jsvc
28308 admin 15 0 1436M 1.4G 40420 S 0.3 14.6 0:01 1 jsvc
So you can see, I have 6 process at the top of my "top" session that are in a run state and is using 20% of cpu each. In 2 or 3 hours, i'll most likely get another one. This server has 4 cpus on it.
I'm sure a restart will clear it but I'd like to know what's causing it. I have the lambda probe installed but can't match the PID from "top" to anything in the thread list for me to "kill" off.
Any ideas?
RE JSVC Process
did you ever get any information on this. We are seeing the same issue with LP 4.2 in pre prod environment. We only have one jsvc process for the tomcat-cp which is eating 99% of the cpu.
A restart of the tomcat resolves the issue (for a time). Not a huge problem at the moment as it is pre prod - but students will be back next month and it makes me a little wary of what might happen to the production portal when all the students start hitting it.
Thanks,
James
jsvc process
We are also seeing the same randomly did anyone find out what was causing the services to hang.
Thanks
Billy
Same here
Just bumping this as we've noticed the exact same thing happening here. jsvc running around 20%, can't figure out what is causing it.
Was anyone able to find a solution for this?
JSVC Process and CPU
We are running Luminis 4.2.1.78 on RHEL 5 (yes, I said 5)... but have experienced this since the 4.1.x versions...
You can see my earlier comment in this thread which states what our problem was ...
What we finally found with the help of Sungard support is that there may bewebapps in the server.xml which are may not be being used which then should be commented out of the $CP_ROOT/products/tomcat/tomcat-cp/conf/server.xml file . . . in our case we were not using the following - after commenting them out and restarting the cp-webserver, the problem disappeared..
<Context path="/pacwa"
crossContext="true"
docBase="/xxx/xxx/webapps/pacwa"/>
<Context path="/spml"
crossContext="true"
docBase="/xxx/xxx/webapps/spml"/>
<Context path="/ws"
crossContext="true"
docBase="/xxx/xxx/webapps/webservice"/>
You should also take the chance to look at the tuning of your system at the same time - FYI our SR with Sungard should have the information in it - SR 1-556519963.
Hopefully this will assist someone having the same issues.
James