@prunkdump:
I've also reported the bug also on Canonical's internal (subscription based) 
support portal before I knew it had been reported on launchpad as well.

Ritesh told me earlier to run a special debug version of cupsd under
valgrind. The logrotate configuration for cupsd did not work well with
valgrind, so I disabled logrotate for cupsd. I than ran cupsd under
valgrind for some time, and the crashes were gone.

To make sure it wasn't valgrind that caused the bug to disappear, I then
ran the debug version of cupsd without valgrind, but forgot to reenable
logrotate. cupsd still did not crash.

Then I noticed that logrotate was disabled. I reenabled logrotate, and
soon after, cupsd showed the regular crashes.

I subsequently identified a web application using Java printing API to
always be involved in the crashes (i.e. a number of requests from the
machine showing in the logs just before the crashes). A couple of days
ago, I redirected that web application to a testing cups instance to try
to reproduce the problem there.

The production instance hasn't crashed since.

The testing instance hasn't crashed either, but the crashes usually only
occur every couple of days. So they might still come.

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/512220

Title:
  cupsd crash "Closing on unknown HTTP state 0"

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/cups/+bug/512220/+subscriptions

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to