cups-polld hung

angelb at bugarin.us angelb at bugarin.us
Fri Sep 10 07:42:58 PDT 2004


I have a Solaris client running CUPS 1.1.18 and on occasion, it would appear to be hung. When you do an lpstat -a, no printer queue defined
is returned. The problem appears to correlate with the cups-polld, poll process for the CUPS server whose IP is 10.208.12.65, just accumulates process time. Please see the following:

# lpstat -a
#

# ps -ef|grep -i cups
cups   857   834  0   Aug 19 ?        0:01 cups-polld 10.208.12.67 631 30 631
root   834     1  0   Aug 19 ?        0:08 /opt/TWWfsw/cups1118/sbin/cupsd -c /etc/opt/TWWfsw/cups1118/cupsd.conf
cups   858   834 25   Aug 19 ?       15566:58 cups-polld 10.208.12.65 631 30 631

While the quick fix to this is refresh the cupsd daemon, it's not a viable solution to a production system. I have implemented a cron process to check this process every 2 minutes, and will restart the cupsd if needed, to minimize queue backlogs.

Has anyone seen this happened before and if so, what was the cause and fix?

This appears to happen only on Solaris systems. I'm also running CUPS clients on AIX and so far not one of them hung.

As I mentioned above, the Solaris client is running CUPS 1.1.18 and my CUPS server is 1.1.20 in RHEL.

Thanks!
Angel









More information about the cups mailing list