logcheck package updated: Now hourly 'system check' mails don't come

[Trane Francks]

Running WHM 11.38.1.15. A couple of days ago, upcp updated/installed a handful of packages:

[20130815.150033] [25330] Updating:
[20130815.150033] [25330] logcheck noarch 1.3.13-6.el5 epel 116 k
[20130815.150033] [25330] Installing for dependencies:
[20130815.150033] [25330] liblockfile x86_64 1.08-9.el5 epel 19 k
[20130815.150033] [25330] lockfile-progs x86_64 0.1.15-2.el5 epel 19 k
[20130815.150033] [25330] perl-IPC-Signal noarch 1.00-2.el5 epel 7.6 k
[20130815.150033] [25330] perl-Proc-WaitStat noarch 1.00-2.el5 epel 8.4 k
[20130815.150033] [25330] perl-mime-construct noarch 1.11-2.el5 epel 21 k

Subsequent to the update, the system check mails were borked. These came instead:

Subject: Cron <[email protected]> if [ -x /usr/sbin/logcheck ]; then nice -n10 /usr/sbin/logcheck; fi
Body: /usr/sbin/logcheck: line 100: kill: (28211) - No such process

With the process number changing with every hour that ticked by. Being that I know just enough to be dangerous, I tried to fix access rights by:

setfacl -m u:logcheck:r /var/log/messages
setfacl -m u:logcheck:r /var/log/secure

That got rid of the error messages, but now nothing comes at all. Nothing whatsoever.

Help me Obi-Wan Kenobi. You're my only hope.

 

[Trane Francks]

Quick update. Now the original cron errors are back.

No ideas, anybody?

 

[cPanelMichael]

Hello

Logwatch is a system package that is not managed by cPanel. I do notice discussion of this error message at:

https://bugzilla.redhat.com/show_bug.cgi?id=678436

Thank you.

 

[Trane Francks]

Thanks, Michael.

I guess it's just one of those mysteries of open source where something that has worked perfectly well for years suddenly gets broken by an update. I seem to have fixed it, but we'll only know for sure once the next logrotate happens.

Cheers.