[cups.general] CUPS stops and restarts jobs

Paul Mitchell pmitchel at email.unc.edu
Tue Feb 14 12:34:42 PST 2012


On Tue, 14 Feb 2012, Michael Sweet wrote:

> The timing of the issue is suspicious - Red Hat uses logrotated to 
> rotate the CUPS logs (instead of letting CUPS do it). Do you see any 
> "shutting down" or "restart" messages in the log?

Hello Michael,
    You make a very good point; I don't see any restart or logrotated 
messages in the logs, OTOH, the problem occur at 08:03 and, curiously, we 
manage CPS via puppet and at 08:03 we get:

Feb 14 08:03:47 ccpapp04 puppetd[9781]: 
(//peoplesoft::cups/File[/etc/cups/ppd/our142.ppd]/checksum) checksum 
changed '{md5}9a1224ae47894ce25bb11960e5d55e1b' to 
'{md5}a5c0e04ac957ba1507ae37ba4a714e06'
Feb 14 08:03:48 ccpapp04 puppetd[9781]: 
(//peoplesoft::cups/File[/etc/cups/ppd/our142.ppd]/content) content 
changed '{md5}a5c0e04ac957ba1507ae37ba4a714e06' to 
'{md5}9a1224ae47894ce25bb11960e5d55e1b'
Feb 14 08:03:48 ccpapp04 puppetd[9781]: 
(//peoplesoft::cups/File[/etc/cups/ppd/our142.ppd]/mode) mode changed 
'644' to '755'
Feb 14 08:03:48 ccpapp04 puppetd[9781]: 
(//peoplesoft::cups/File[/etc/cups/printers.conf]/checksum) checksum 
changed '{md5}10fc065bd62b3a2b1a6d1bc0dda66a08' to 
'{md5}29bd8f0e4d20cee026bdd882578c950c'
Feb 14 08:03:48 ccpapp04 puppetd[9781]: 
(//peoplesoft::cups/File[/etc/cups/printers.conf]/content) content changed 
'{md5}29bd8f0e4d20cee026bdd882578c950c' to 
'{md5}10fc065bd62b3a2b1a6d1bc0dda66a08'
Feb 14 08:03:48 ccpapp04 setroubleshoot: SELinux is preventing cupsd 
(cupsd_t) "rename" to ./printers.conf (cupsd_etc_t). For complete SELinux 
messages. run sealert -l a27d6de3-64ba-4bce-a4cc-8f119a4ec7e8
..
..
..
(//peoplesoft::cups/Exec[cups-our142-enable]/returns) executed 
successfully
Feb 14 08:03:50 ccpapp04 puppetd[9781]: 
(//peoplesoft::cups/Exec[cups-our142-enable]) Triggering 'refresh' from 2 
dependencies

hmm, what if puppet refreshs printer while the printer is operating?

> As for restarting from the stop point, CUPS does not support that 
> (really hard to do reliably with arbitrary printers/drivers/interfaces) 
> so let's focus on preventing the restart in the first place.

I understand that the information required for this is beyond the 
capabilities of the software involved.  I think I'll look at the puppet 
configuration on the puppet master.

Thanks,

Paul Mitchell

==============================================================================
         Paul Mitchell
 	Enterprise Systems
         email: pmitchel at email.unc.edu
 	location: 440 Franklin, cubby 1213
 	desk phone: 919 962-2521 (Is here!^)
==============================================================================







More information about the cups mailing list