cl-selector bug with Cloud Linux 6.7 and Plesk 12.0.18

[cloudlinux]

unhandled python exception in /usr/bin/cl-selector

[george]

When trying to add a new user to CageFS via the Plesk extension, I get the following errors:

Sep 21 16:37:28 host105 abrt: detected unhandled Python exception in /usr/bin/cl-selector
Sep 21 16:37:28 host105 abrt-server[86859]: Saved Python crash dump of pid 86858 to /var/spool/abrt/pyhook-2015-09-21-16:37:28-86858
Sep 21 16:37:28 host105 abrtd: Directory pyhook-2015-09-21-16:37:28-86858 creation detected
Sep 21 16:37:29 host105 abrtd: sosreport run failed with exit code 1, log follows:
Sep 21 16:37:29 host105 abrtd: a non-existing plugin (boot) was specified in the command line
Sep 21 16:37:29 host105 abrtd: sosreport (version 2.2)
Sep 21 16:37:29 host105 abrtd: post-create on /var/spool/abrt/pyhook-2015-09-21-16:37:28-86858 exited with 1
Sep 21 16:37:29 host105 abrtd: Deleting problem directory /var/spool/abrt/pyhook-2015-09-21-16:37:28-86858
Sep 21 16:37:30 host105 abrt: detected unhandled Python exception in /usr/bin/cl-selector
Sep 21 16:37:30 host105 abrt-server[86885]: Not saving repeating crash in /usr/bin/cl-selector

This problem seems to only affect adding new users to CageFS, existing users work just fine.

Please point me in the right direction for resolving this problem.

I already made sure that the latest yum updates have been applied.

Thank you in advance for your assistance.

- George

[bogdan.sh]

That is quite interesting we never saw, but would like to investigate. No ideas what is going on yet, could you please create a support ticket so we could check the issue in place?

[george]

Thank you for following up Bogdan. Ticket number QAV-923-43977 has been submitted and resolved. While not the ideal fix, restarting cageFS solved the problem. I was asked to submit another ticket if the problem recurs, and I will definitely do so if necessary.

In closing, thank you for your EXCELLENT product and FANTASTIC service!

[bogdan.sh]

Well, we would really like to find out the core of the issue, next time issue happens and abrt will save a report please store it in some safe location.

[george]

I looked for abrt output, but the report directory referenced in the following entries in the message log does not exist. Do I have something misconfigured, perhaps? Do I need to do something special to save the output from abrt? The abrt daemon indicates that it deleted the report folder.

Here are the entries from the message log

Sep 21 15:45:36 host105 abrt: detected unhandled Python exception in /usr/bin/cl-selector
Sep 21 15:45:36 host105 abrt-server[1043771]: Saved Python crash dump of pid 1043768 to /var/spool/abrt/pyhook-2015-09-21-15:45:36-1043768
Sep 21 15:45:36 host105 abrtd: Directory pyhook-2015-09-21-15:45:36-1043768 creation detected
Sep 21 15:45:37 host105 abrtd: sosreport run failed with exit code 1, log follows:
Sep 21 15:45:37 host105 abrtd: a non-existing plugin (boot) was specified in the command line
Sep 21 15:45:37 host105 abrtd: sosreport (version 2.2)
Sep 21 15:45:37 host105 abrtd: post-create on /var/spool/abrt/pyhook-2015-09-21-15:45:36-1043768 exited with 1
Sep 21 15:45:37 host105 abrtd: Deleting problem directory /var/spool/abrt/pyhook-2015-09-21-15:45:36-1043768
Sep 21 15:45:41 host105 abrt: detected unhandled Python exception in /usr/bin/cl-selector
Sep 21 15:45:41 host105 abrt-server[1043803]: Not saving repeating crash in /usr/bin/cl-selector

And the abrt directory does not have any reports

$# ls -lAF /var/spool/abrt/
total 16
drwxr-xr-x 2 abrt abrt 4096 Sep 21 16:37 ./
drwxr-xr-x 14 root root 4096 Feb 18 2015 ../
-rw-r--r-- 1 root root 0 Feb 18 2015 abrt-db
-rw------- 1 root root 30 Sep 12 15:45 last-ccpp
-rw------- 1 root root 20 Sep 21 16:37 last-via-server

I can give you ssh access to the server if that will help with your diagnosis.

Thank you for your help Bogdan.

[bogdan.sh]

Not sure about abrt and how to configure it right. But did you tried solution I provide you in ticket? For all, there was a bugreport that cagefs hooks are duplicated in Plesk, issue has been fixed in latest packages in beta, updating to beta should fix it.