defunct all over the place

[denisdekat09]

I just noticed that I am building up loads of defunct processes on my servers, particularly RH9 and the bwlimit ones (had one with a bunch of exim also):

root@server01 [/usr/local]# ps auxw | grep defunct
root 6526 0.0 0.0 0 0 ? ZN Sep15 0:00 [upcp <defunct>]
root 31881 0.0 0.0 0 0 ? ZN Sep15 0:00 [bwlimit <defunct>]
root 9603 0.0 0.0 0 0 ? ZN Sep16 0:00 [bwlimit <defunct>]
root 8747 0.0 0.0 0 0 ? ZN Sep16 0:00 [bwlimit <defunct>]
root 22489 0.0 0.0 0 0 ? ZN Sep16 0:00 [bwlimit <defunct>]
root 16141 0.0 0.0 0 0 ? ZN Sep16 0:00 [bwlimit <defunct>]
root 31898 0.0 0.0 0 0 ? ZN Sep16 0:00 [bwlimit <defunct>]
root 5912 0.0 0.0 0 0 ? ZN Sep16 0:00 [bwlimit <defunct>]
root 395 0.0 0.0 0 0 ? ZN Sep16 0:00 [bwlimit <defunct>]
root 9248 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 9249 0.0 0.0 0 0 ? ZN Sep17 0:00 [reachedbwlimit <defunct>]
root 15921 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 12227 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 23888 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 9475 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 886 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 10888 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 22642 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 11575 0.0 0.0 0 0 ? ZN Sep17 0:00 [bwlimit <defunct>]
root 21647 0.0 0.0 0 0 ? ZN Sep18 0:00 [bwlimit <defunct>]
root 24566 0.0 0.0 0 0 ? ZN Sep18 0:00 [bwlimit <defunct>]
root 25860 0.0 0.0 0 0 ? ZN Sep18 0:00 [bwlimit <defunct>]
root 28491 0.0 0.0 0 0 ? ZN Sep18 0:00 [bwlimit <defunct>]
root 2929 0.0 0.0 0 0 ? ZN Sep18 0:00 [bwlimit <defunct>]
root 12675 0.0 0.0 0 0 ? ZN Sep18 0:00 [bwlimit <defunct>]
root 16346 0.0 0.0 0 0 ? ZN Sep18 0:00 [bwlimit <defunct>]
root 4588 0.0 0.0 0 0 ? ZN 04:11 0:00 [bwlimit <defunct>]
root 9506 0.0 0.0 0 0 ? ZN 10:14 0:00 [bwlimit <defunct>]
root 21239 0.0 0.0 0 0 ? ZN 16:15 0:00 [bwlimit <defunct>]
mailnull 21527 0.1 0.0 0 0 ? ZN 18:59 0:00 [exim <defunct>]
root 26368 0.0 0.0 4456 640 pts/1 S 19:00 0:00 grep defunct

I tried

ps auwx | grep defunct | awk '{fs=" "}{print $2}'| xargs kill -9

But nothing kills them Treid rebooting one server and it cleared them up, but man I hate rebooting everything all over the place. Any ideas why these could be happening?

Thanks

[AbeFroman]

Something must be stopping those mid process, do you have a process kill or memory limiter?

[denisdekat09]

nope, not at all I monitor these guys, but they do not have anythng killing processes that I ever put in...

[denisdekat09]

This is coming back on this server.. I guess I will be opening a ticket

[AbeFroman]

What you find out?

[seeyes]

I boys,
I have the same problem... there are more processo like:
[bwlimit] <defunct>
from the last server restart. When I have restat the server, I don't see any [bwlimit] process, but now (after two days) I have 195 process...

I use Fedora Core 2 on a WHM 9.9.0 cPanel 9.9.0-E56

Thank you

[denisdekat09]

sorry, I thought they went away after reboot, but they are coming back. I will open a ticket for sure this time

I have this on RH9 and RH7.3 happening so far...

I don't have as many as before:

ps auxw | grep defunct
root 9920 0.0 0.0 0 0 ? ZN Sep19 0:00 [bwlimit <defunct>]
root 16222 0.0 0.0 0 0 ? ZN Sep20 0:00 [bwlimit <defunct>]
root 26186 0.0 0.0 0 0 ? ZN Sep20 0:00 [bwlimit <defunct>]
root 32695 0.0 0.0 0 0 ? ZN Sep20 0:00 [bwlimit <defunct>]
root 28139 0.0 0.0 0 0 ? ZN Sep20 0:00 [bwlimit <defunct>]
root 5902 0.0 0.0 0 0 ? ZN Sep21 0:00 [bwlimit <defunct>]
root 8343 0.0 0.0 0 0 ? ZN Sep21 0:00 [bwlimit <defunct>]


But still get them and they refuse to die

I tried to use kill -9 on these but no luck...

[Promethyl]

This also happens to me on RHE with Cpanel current.

[mydomain]

Same issue for us also with bwlimit defunct processes

[Promethyl]

GC doesn't clean these until the owner process finishes. I sighup'd mine, and the defunts were removed, but ... who's to say it was coincidence.

[dhost]

Same here more than 79 processes. /scripts/upcp killed all except 1

[sasha]

Same problem here
WHM 9.9.0 cPanel 9.9.0-C55
Fedora i686 - WHM X v3.1.0


/etc/init.d/cpanel restart
makes zombies go away

[haze]

One of you should drop in a bug report so that the dev team knows this problem exists:
http://bugzilla.cpanel.net

[webits]

I'm having the same problem anyone report a bug if so whats the link so i can add on there.

[SupermanInNY]

Same problem here
WHM 9.9.0 cPanel 9.9.0-C55
Fedora i686 - WHM X v3.1.0


/etc/init.d/cpanel restart
makes zombies go away

WHM 9.9.0 cPanel 9.9.1-S50
RedHat 9 i686 - WHM X v3.1.0

/etc/init.d/cpanel restart

Did the trick for me!


-Alon.