http://bugzilla.novell.com/show_bug.cgi?id=622058 http://bugzilla.novell.com/show_bug.cgi?id=622058#c1 Johannes Meixner <jsmeix@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P3 - Medium Status|NEW |ASSIGNED Platform|Other |All OS/Version|Other |openSUSE 11.3 --- Comment #1 from Johannes Meixner <jsmeix@novell.com> 2010-07-14 07:45:24 UTC --- I cannot reproduce it on my machines because here the cupsd finish sufficiently fast. Perhaps it happens only when the cupsd is under high load? Nevertheless you are right: "rccups stop" sends SIGTERM to the cupsd so that it can do its clean up work before it actually finishes. While cupsd does its clean up work the subsequent "rccups start" runs "startproc /usr/sbin/cupsd" which starts /usr/sbin/cupsd only if no cupsd process is found but there could be still the old cupsd running which still does its clean up work until it actually finishes so that in the end the old cupsd finishes correctly but no new cupsd was started. As a workaround you may use: rccups stop && sleep 3 && rccups start -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=622058 http://bugzilla.novell.com/show_bug.cgi?id=622058#c3 Johannes Meixner <jsmeix@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO InfoProvider| |jnelson-suse@jamponi.net --- Comment #3 from Johannes Meixner <jsmeix@novell.com> 2010-07-14 08:05:42 UTC --- Please provide feedback if attachment #375799 works o.k. for you. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=622058 http://bugzilla.novell.com/show_bug.cgi?id=622058#c5 Johannes Meixner <jsmeix@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO CC| |mt@novell.com, | |werner@novell.com InfoProvider| |werner@novell.com --- Comment #5 from Johannes Meixner <jsmeix@novell.com> 2010-07-15 10:21:22 UTC --- I think I found the root cause why it never happened before openSUSE 11.3, see bug #595796. As far as I see before bug #595796 was fixed, "rccups stop" did "killproc -TERM /usr/sbin/cupsd" and killproc did send SIGTERM, then waits by default up to 5 seconds and then sends a SIGKILL so that in this 5 seconds the cupsd could do its clean up work and finish. Since bug #595796 was fixed for openSUSE 11.3, killproc with an explicite signal specified like "killproc -TERM" sends only SIGTERM and exits immediately. If my analysis is right, I wonder if I should move the 10 seconds wait until cupsd had acually finished from the specific "rccups restart" case to the general "rccups stop" case. The general question is what is expected by the user (or by LSB) when "rc<whatever> stop" is called. Should this only trigger the service to be stoped or should (or must) this also wait until the service had actually stopped? In other words: What is expected when the user calls rc<whatever> stop ; rc<whatever> start If this must be equivalent to "rc<whatever> restart" then "rc<whatever> stop" must wait until the service had actually finished. I have in particular a system shutdown in mind where many "rc<whatever> stop" are executed. If each of them waits until each service had actually stopped, the services may not finish simultaneously but only one by one which may delay the system shutdown needlessly. Or are the many "rc<whatever> stop" already called as separated processes so that they run simultaneously in any case? Werner, could you provide some background information? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=622058 http://bugzilla.novell.com/show_bug.cgi?id=622058#c7 Johannes Meixner <jsmeix@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #375799|0 |1 is obsolete| | --- Comment #7 from Johannes Meixner <jsmeix@novell.com> 2010-07-15 13:10:31 UTC --- Created an attachment (id=376166) --> (http://bugzilla.novell.com/attachment.cgi?id=376166) Better fix for /etc/init.d/cups -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=622058 http://bugzilla.novell.com/show_bug.cgi?id=622058#c9 Johannes Meixner <jsmeix@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |RESOLVED InfoProvider|jnelson-suse@jamponi.net | Resolution| |FIXED Target Milestone|--- |Factory --- Comment #9 from Johannes Meixner <jsmeix@novell.com> 2010-07-15 13:43:54 UTC --- I submitted CUPS with /etc/init.d/cups as in attachment #376166 to the Printing project via submitrequest 42968 and to openSUSE:Factory via submitrequest 42970 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=622058 http://bugzilla.novell.com/show_bug.cgi?id=622058#c10 --- Comment #10 from Bernhard Wiedemann <bwiedemann@suse.com> --- This is an autogenerated message for OBS integration: This bug (622058) was mentioned in https://build.opensuse.org/request/show/42970 Factory / cups -- You are receiving this mail because: You are on the CC list for the bug.