Peter Much wrote: > Hi all, > > I have a functional problem with the director daemon. > > The problem depends on the configuration of the scheduler. As long as > there are only schedules in effect with any granularity of *daily*, > there is no problem. > > But as soon as you configure a schedule with any kind of granularity > of *hourly* (that is, a schedule that is run at specific times every > hour) *AND* put that schedule into effect by calling it from some > job, then following problem appears: > > The SIGTERM signal doesn't work anymore. Instead of terminating, the > bacula-dir goes into an endless loop and eats all available cpu. > Sending *another* SIGTERM at that point will terminate the daemon > correctly. > > The problem therein is: when doing a main system shutdown, the > rc.shutdown will send _one_ SIGTERM to the daemon, and then will > just wait for the daemon to terminate. > It will wait for 15 minutes; then a watchdog hits in and does a > forced halt of the whole system - which isnt nice for all the other > innocent daemons running. > > It is not too troublesome, as we usually do the shutdowning only > twice a year, but it should occasionally be fixed, just for beauty.
Please provide a very specific example. I know you gave a general example, but a specific example allows others to test the same thing and see if this is a problem that can be reproduced on multiple platforms. -- Dan Langille - http://www.langille.org/ BSDCan - The Technical BSD Conference: http://www.bsdcan.org/ PGCon - The PostgreSQL Conference: http://www.pgcon.org/ ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
