Yep that was a fresh jail.
All the pid’s of the “c++ processes” are shown when I run pgrep -g 'cat /var/run/plex/plex.pid'
Actually took me a bit to find the flag I wanted but the man page helped 
ps auxd
lmao man always works
It seems like the c++ processes are being spawned independently somehow but are associating PMS as its parent when the PMS service is stopped?
USER PID %CPU %MEM VSZ RSS TT STAT STARTED TIME COMMAND
plex 8613 30.5 0.0 440 352 - RJ 03:53 3:36.78 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpVuQGV9 -luuid
plex 8861 28.9 0.0 2972 2508 - RJ 03:56 1:49.45 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpKyofqt -luuid
plex 8858 28.6 0.0 440 352 - RJ 03:56 1:48.09 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpC6Qxvb -luuid
plex 8855 28.4 0.0 2972 2508 - RJ 03:56 1:48.50 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpFmW4U8 -luuid
plex 8859 28.2 0.0 2972 2452 - RJ 03:56 1:48.30 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpyGl8nj -luuid
plex 8490 28.1 0.0 440 352 - RJ 03:47 9:33.77 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpqvqgni -luuid
plex 8615 27.7 0.0 440 352 - RJ 03:53 3:35.55 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpqbTxg3 -luuid
plex 8609 27.4 0.0 300 192 - RJ 03:53 3:37.04 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpPpgx84 -luuid
plex 8614 26.6 0.0 440 336 - RJ 03:53 3:35.41 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpPavY2a -luuid
plex 8544 26.2 0.0 7068 2668 - RJ 03:50 6:32.02 /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpdhszjx -luuid
plex 8487 0.2 0.4 137224 73120 - IsJ 03:47 0:02.00 /usr/local/share/plexmediaserver/Plex_Media_Server (Plex Media Server)
plex 8542 0.0 0.1 27788 20432 - IJ 03:50 0:00.05 - /usr/local/share/plexmediaserver/Plex Tuner Service /usr/local/share/plexmediaserver/Resources/Tuner/Private /usr/local/share/plexmediaserver/Resources/Tuner/Shared 1.19.1.2701-
plex 8904 0.0 0.2 37128 30416 - IJ 03:59 0:01.02 - Plex Plug-in [com.plexapp.agents.none] /usr/local/share/plexmediaserver/Resources/Plug-ins-6327e27bf/Framework.bundle/Contents/Resources/Versions/2/Python/bootstrap.py --server-
plex 8912 28.2 0.0 440 352 - RJ 03:59 0:38.82 `-- /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmpv58GkU -luuid
plex 8905 0.0 0.2 37128 30416 - IJ 03:59 0:01.02 - Plex Plug-in [com.plexapp.agents.lyricfind] /usr/local/share/plexmediaserver/Resources/Plug-ins-6327e27bf/Framework.bundle/Contents/Resources/Versions/2/Python/bootstrap.py --se
plex 8909 29.7 0.0 7068 2668 - RJ 03:59 0:38.94 `-- /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmppODwjy -luuid
plex 8906 0.0 0.2 37128 30416 - IJ 03:59 0:01.03 - Plex Plug-in [com.plexapp.system] /usr/local/share/plexmediaserver/Resources/Plug-ins-6327e27bf/Framework.bundle/Contents/Resources/Versions/2/Python/bootstrap.py --server-versi
plex 8915 28.5 0.0 440 352 - RJ 03:59 0:39.11 `-- /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmposc9Zv -luuid
plex 8907 0.0 0.2 37128 30416 - IJ 03:59 0:01.02 - Plex Plug-in [com.plexapp.agents.imdb] /usr/local/share/plexmediaserver/Resources/Plug-ins-6327e27bf/Framework.bundle/Contents/Resources/Versions/2/Python/bootstrap.py --server-
plex 8913 27.6 0.0 132 16 - RJ 03:59 0:39.02 `-- /bin/sh /usr/bin/c++ -Wl,-t -o /tmp/tmp6UyTLd -luuid
root 7226 0.0 0.0 6420 2460 - SsJ 03:16 0:00.00 /usr/sbin/syslogd -c -ss
root 7282 0.0 0.0 6460 2340 - IsJ 03:16 0:00.01 /usr/sbin/cron -J 15 -s
root 8470 0.0 0.0 6948 2852 0 IJ 03:47 0:00.01 login [pam] (login)
root 8471 0.0 0.0 7488 3972 0 SJ 03:47 0:00.01 - -csh (csh)
root 9005 0.0 0.0 7000 2828 0 R+J 04:01 0:00.00 `-- ps auxd
Also it looks like more have spawned
Nope its not on there own its Python! need to look a bit closer but its the python stuff (Plex Script Host)
Ahhh I didn’t see the ones under the Plex Plug-In branch my bad
Ok still no clue how this would happen but its a start, I’ll see if this makes sense to the PMS devs and update you when I have more info.
Ok! Thanks for all your help btw
So mikec and me were PM’ing each other to diagnose stuff, and what we found was this:
Several instances of sh were being spawned running some c++ after the core component failure page was pushed. Afterwards, more sh instances would continually be spawned forever which caused a resource bottleneck. It was hypothesized that this was ultimately a python issue, and would be looked into.
The fix that I did was just use FreeBSD instead of FreeNAS. Everything works well now. I have a feeling that this is ultimately a FreeNAS problem, since when FreeNAS was installed, for some reason doing things like installing packages and fetching the port tree would take a really long time (like greater than an hour for fetching the port tree). After installing FreeBSD, everything takes a more reasonable amount of time. So I think that this wasn’t actually Plex’s problem, but rather something wrong with FreeNAS or something, and Plex was just timing out as it should.
1 Like
I have been having the exact same issue you are describing. Failing install, cpu usage spikes. Have not gone into lower level as you have, but looks like same reaction. Is there any way to resolve this? Should I just shelve the idea of using Plex until a new version of FreeNAS or Plex comes out? I have a separate thread as I did not find this one. https://forums.plex.tv/t/freenas-plex-plugin-install-failing/598623
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.
