Zombie processes on my DEA machines.

90 views
Skip to first unread message

Guruprakash S

unread,
Apr 28, 2015, 3:39:36 PM4/28/15
to vcap...@cloudfoundry.org
Hi Team,

On all of my 5 DEA machines, I see a couple of Zombie processes running. We are running on Cloud Foundry v197. This is happening since CF upgrade from v172 to v197 yesterday. Please advise.

vcap@0ea9719f-0524-408f-845f-3d1377272f3d:~$ ps aux | grep 'Z'
USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
root      1189  0.0  0.0      0     0 ?        Z<   17:55   0:00 [dea_logging_age] <defunct>
root      1191  0.0  0.0      0     0 ?        Z<   17:55   0:00 [dea_logging_age] <defunct>
root      2113  0.0  0.0      0     0 ?        Z<   17:55   0:00 [metron_agent_ct] <defunct>
root      2115  0.0  0.0      0     0 ?        Z<   17:55   0:00 [metron_agent_ct] <defunct>


Thanks,
Guruprakash

Erik Jasiak

unread,
Apr 28, 2015, 4:15:15 PM4/28/15
to vcap...@cloudfoundry.org
Hi Guruprakash,
   I'll poke at this more with the loggregator team - all of those zombie processes are logging and metrics related.  I'm still relatively new to CF though, so I'm not sure what parent process is spawning the DEA Logging Agent and Metron Agent on a DEA CF box that isn't reaping the zombies in a timely fashion (hopefully someone else will chime in in a moment.)

  We've also done quite a bit to reduce our resource leaks between CF 197 and CF 201/203 specifically [1][2] , which are frequently related to zombie processes. 

Thanks,
Erik


--
You received this message because you are subscribed to the Google Groups "Cloud Foundry Developers" group.
To view this discussion on the web visit https://groups.google.com/a/cloudfoundry.org/d/msgid/vcap-dev/8e1aed12-759d-4626-975d-17589c3e7b22%40cloudfoundry.org.

To unsubscribe from this group and stop receiving emails from it, send an email to vcap-dev+u...@cloudfoundry.org.

Guruprakash S

unread,
Apr 28, 2015, 6:17:49 PM4/28/15
to vcap...@cloudfoundry.org
Hi Erik,

Did check further and found that the below parent processes are responsible for the zombie process. Thanks for checking this. Will wait on loggregator team for more info.

deaagent (parent process) ---> dea_logging_age (Zombie process)
metron (parent process) ----> metron_agent_ct (Zombie process)


vcap@0ea9719f-0524-408f-845f-3d1377272f3d:~$ pstree -p -s 1189
init(1)───deaagent(1187)───dea_logging_age(1189)

vcap@0ea9719f-0524-408f-845f-3d1377272f3d:~$ pstree -p -s 2113
init(1)───metron(2111)───metron_agent_ct(2113)


Thanks,
Guruprakash
Reply all
Reply to author
Forward
0 new messages