My every-15-minutes job is getting a few "Duplicate job not allowed." messages when things get a bit banked up during some big nighttime jobs. Those messages aren't a problem as there is very little production data backed up at that time and I'd rather not see them.
Looking at the code, such a message is M_FATAL, but even if I suppressed that the job-was-cancelled error message is separate and doesn't record that the reason it was cancelled was because of the duplicate error message: 22-Aug 12:20 bccbacula-dir JobId 10628: Fatal error: JobId 10627 already running. Duplicate job not allowed. 22-Aug 12:20 bccbacula-dir JobId 10628: Error: Bacula bccbacula-dir 5.0.2 (28Apr10): 22-Aug-2010 12:20:50 Build OS: x86_64-pc-linux-gnu debian squeeze/sid ... So I guess I'm going to have to put up with it in the short term, but would you been open to introducing the concept of messages ID's where every message (or at least most messages) gets a unique ID (in addition to the INFO/WARN/ERROR/FATAL class it already has)? eg message ID 1337 might be a M_FATAL error meaning "Job already running. Duplicate job not allowed.". This has several advantages to what is done now with ad-hoc message text: . First message of the highest class could be stored against the Job record for things like bat and nagios to report on, eg if the job ended in a WARN state then the first M_WARN message ID would be recorded. . Much finer grained reporting, eg I want all M_INFO except 'Will not descend' messages and all M_FATAL except 'Duplicate job not allowed' messages . Consistent message text - the "JobId %d already running. Duplicate job not allowed." message is in the code in two places . Makes internationalisation easier - the message text is stored in a big array making it easy to have different language versions of the text (not sure if this is an attraction or not...) Comments? Thanks James ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ Bacula-devel mailing list Bacula-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-devel
