Hello,
Sorry, but this is too complicated for me. I am very overloaded so need
questions/output short and simple. Perhaps someone else can help you with the
questions you have.
What I would like to see is: the output showing how job messages are
intermixed. I don't need the whole log, but if you clip stuff out here and
there, it won't help me understand the problem.
On Wednesday 29 March 2006 16:00, Heinz Ulrich Stille wrote:
> On Wednesday 29 March 2006 14:09, you wrote:
> > > I've got three jobs with different priorities in each backup, so
> > > if one backup isn't able to run (perhaps tape wasn't changed, or
> > > I'm manually doing something else) I'll get the success messages
> > > of the next run interleaved with the errors of the timed out one.
> >
> > Can you show us an example? Some of what you say concerning why the job
> > exceeds the Max Start Delay doesn't sound pertinent.
>
> I've got three jobs:
> "full": the main system backup, priority 10
> "catalog": bacula catalog backup, priority 50
> "closing": admin job to eject the tape, priority 99
>
> All Jobs have "Max Start Delay = 22 hours" and the following schedule:
> Schedule { Name = "Weekly"; Run = Full Pool=Weekly fri at 15:30; Run =
> Incremental Pool=Daily mon-thu at 17:30 }
>
> I started "full" with Level=Full manually on monday. This took until
> tuesday, 17:34 to finish. In the meantime the regular jobs for monday
> and tuesday got scheduled. On monday at 17:30 the jobs for that day
> were run and changed status to "waiting for other / higher priority
> job to finish". When the manual backup finished, there were two instances
> of each job waiting, which got "run" in order of priority. So at 17:34,
> the "full" of monday was failed and the one of tuesday run. This took
> until 19:16. At that time the next priority job, "catalog" from monday
> was examined and failed, then the one from tuesday run. Same for
> "closing" a few minutes later.
>
> The resulting log (director status) is:
> Terminated Jobs:
> JobId Level Files Bytes Status Finished Name
> ========================================================================
> 3 Full 803,049 172,003,756,473 OK 28-Mar-06 17:34 full
> 4 Incr 0 0 Error 28-Mar-06 17:34 full
> 7 Incr 9,788 5,295,056,880 OK 28-Mar-06 19:16 full
> 5 Full 0 0 Error 28-Mar-06 19:16 catalog
> 8 Full 1 101,405,753 OK 28-Mar-06 19:18 catalog
> 6 0 0 Error 28-Mar-06 19:18 closing
> 9 0 0 OK 28-Mar-06 19:19 closing
>
> BTW: column formatting is for a maximum of 99,999,999,999 Bytes - not all
> that much for a real server...
>
> The failure notices sent by mail were:
> 28-Mar 17:34 mail-dir: full.2006-03-27_17.30.00 Fatal error: Job canceled
> because max start delay time exceeded.
>
> 28-Mar 19:16 mail-dir: catalog.2006-03-27_17.30.01 Fatal error: Job
> canceled because max start delay time exceeded.
> 28-Mar 19:16 mail-dir: No prior Full backup Job record found.
> 28-Mar 19:16 mail-dir: No prior or suitable Full backup found. Doing FULL
> backup.
>
> 28-Mar 19:18 mail-dir: closing.2006-03-27_17.30.02 Fatal error: Job
> canceled because max start delay time exceeded.
>
> With the OK messages of the other jobs in between. And now that I look at
> it: Why did the second job (at 19:16) first fail and then get upgraded to
> Full level?
>
> Anyway, my gripe is that the jobs were not cancelled at the proper time,
> while still waiting, but only when they finally got a chance to execute.
>
> Another question: I'd prefer to use "Offline On Unmount"; but I think the
> tape is unmounted after each single job, so that it would be ejected
> between my "full" and "catalog" jobs? If that is the case, is there a way
> around it? Jobs running each other with the "Run" directive still produce
> separate jobs, or not?
>
> MfG, Ulrich
--
Best regards,
Kern
(">
/\
V_V
-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
