Orphaned socket/lp's and lpstat's

[Paul Conklin]

it was using the init.d/cups file, looks like that is -hup for the Restart / reload (both of which I know the user tried after we tracked down who did it)  I'll have to ask if they did a stop, I know they didn't explictly kill it, everything they did was via the init.d/cups file.

> We haven't had any new reports of zombies in a long time now; find out how cupsd was restarted - SIGHUP or SIGTERM should be find, SIGKILL will prevent cupsd from cleaning up after itself.
>
> On Sep 30, 2011, at 6:40 AM, Paul Conklin wrote:
>
> > I'm not sure if this is a defect, an enhancement, or just an instance that I need to go start slapping people's hand's with rulers, so I figure I'll start here and log a bug if that's the consensus.
> >
> > CUPS 1.4.7 / RHEL 5.3 and 5.5
> >
> > We started noticing some orphaned lpstat's as well as socket/lp's out on the system and the lpstats were just chewing up CPU in top.  going through the error_log (we leave at info) i noticed that the timestamp on the processes correlates to the exact same time that somebody did a service cups reload or a restart.  Anyone know if this is patched in any other versions or is this a bug as I would think the daemon should either keep track of it's children or the children should "call home" occasionally to make sure they aren't orphans.
> >
> > Thanks in advance.
> > _______________________________________________
> > cups mailing list
> > cups at easysw.com
> > http://lists.easysw.com/mailman/listinfo/cups
>
> __________________________________________________
> Michael Sweet, Senior Printing System Engineer, PWG Chair
>