[cups.general] (1.5.4) cupsd dies while trying to display a list of completed jobs

Alex Korobkin korobkin+cups at gmail.com
Thu Oct 18 14:29:58 PDT 2012


Hi team,

I've recently updated my Ubuntu 10.04 print server to CUPS 1.5.4 and
encountered this problem:

When i try to get a list of completed jobs via web-interface (Jobs > Show
Completed Jobs), server prints page header, waits for about a minute, then
displays nothing below the header.

With help of *top* I see that as soon as i hit "Show Completed Jobs"
button, CUPS starts to use 100% of CPU, then gets killed and respawned with
a new PID. *error_log* shows that CUPS receives job listing request, finds
out it needs to process 749 completed jobs from 53 printers, reads
their attributes one by one, and then suddenly dies as if something has
killed it.

Is there an internal timeout (60 sec?) that would kill cupsd in this
situation?
Is there a way to make CUPS process these jobs page by page in order to
reply faster, instead of going through all those completed jobs?

Here is an excerpt from the error_log (debug2):

== Server starts:
I [18/Oct/2012:19:47:52 +0000] Resuming new connection processing...
<snip>
D [18/Oct/2012:19:47:52 +0000] cupsdSetBusyState: newbusy="Not busy",
busy="Not busy"
<snip>
D [18/Oct/2012:19:47:52 +0000] Discarding unused server-started event...
<snip>
D [18/Oct/2012:19:47:52 +0000] Report: clients=1
D [18/Oct/2012:19:47:52 +0000] Report: jobs=749
D [18/Oct/2012:19:47:52 +0000] Report: jobs-active=0
D [18/Oct/2012:19:47:52 +0000] Report: printers=53
D [18/Oct/2012:19:47:52 +0000] Report: printers-implicit=0
D [18/Oct/2012:19:47:52 +0000] Report: stringpool-string-count=648288
D [18/Oct/2012:19:47:52 +0000] Report: stringpool-alloc-bytes=39360
D [18/Oct/2012:19:47:52 +0000] Report: stringpool-total-bytes=11849752
<snip>

== Server tries to create a report on completed jobs:
d [18/Oct/2012:19:49:48 +0000] get_jobs: job->id=167215,
dest="james-color", username="johndoe", state_value=9, attrs=(nil)
D [18/Oct/2012:19:49:48 +0000] [Job 167215] Loading attributes...
d [18/Oct/2012:19:49:48 +0000] cupsdCheckGroup(username="anonymous",
user=(nil), groupname="OWNER")
d [18/Oct/2012:19:49:48 +0000] get_md5_password(username="anonymous",
group="OWNER", passwd=0x7fff8fb6d3b0)
d [18/Oct/2012:19:49:48 +0000] cupsdCheckGroup(username="anonymous",
user=(nil), groupname="root")
d [18/Oct/2012:19:49:48 +0000] get_md5_password(username="anonymous",
group="root", passwd=0x7fff8fb6d3b0)
d [18/Oct/2012:19:49:48 +0000] cupsdCheckGroup(username="anonymous",
user=(nil), groupname="cups-admins")
d [18/Oct/2012:19:49:48 +0000] get_md5_password(username="anonymous",
group="cups-admins", passwd=0x7fff8fb6d3b0)
d [18/Oct/2012:19:49:48 +0000] copy_attrs(to=0x7f0811633380,
from=0x7f08113dd9b0, ra=0x7f081163dec0, group=2, quickcopy=0)
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113ddcd0[time-at-creation,2,21])
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113ddd10[time-at-processing,2,21])
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113ddd50[time-at-completed,2,21])
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113ddd90[job-id,2,21])
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113dddd0[job-state,2,23])
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113dde10[job-media-sheets-completed,2,21])
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113dde90[job-k-octets,2,21])
d [18/Oct/2012:19:49:48 +0000] copy_attribute(0x7f0811633380,
0x7f08113ddfd0[job-printer-state-message,2,41])
d [18/Oct/2012:19:49:48 +0000] get_jobs: job->id=167216,
dest="james-color", username="joesmith", state_value=9, attrs=(nil)
*D [18/Oct/2012:19:49:48 +0000] [Job 167216] Loading attributes...*
I [18/Oct/2012:19:49:48 +0000] Remote access is enabled.
I [18/Oct/2012:19:49:48 +0000] Loaded configuration file
"/etc/cups/cupsd.conf"
I [18/Oct/2012:19:49:48 +0000] Using default TempDir of
/var/spool/cups/tmp...

Highlighted line is where server suddenly dies and restarts. No indication
why would it die.

Any hints are much appreciated.
-Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cups.org/pipermail/cups/attachments/20121018/0582bb9e/attachment-0001.html>


More information about the cups mailing list