Re: [Bacula-users] Unnecessarily duplicating backups

Hello Ian,

The concept of schedule and start times are the same I guess. One is the one that you set in your schedule resource and the other is the time that the job starts running. Usually, they are slightly different.

Again, if you have duplicated jobs starting and you do not want this situation, maybe you need to review your schedules.

It would help us to understand your issue if you could send here the job and schedule definitions. Also, some outputs and log information are always helpful in these cases.

There is a few possible configurations that can force bacula to run a full instead of a differential or incremental depending on changes in the FileSet. So have a look into your configurations would help us to understand this.

The list of files to be backed up is build when the job starts and not when the job is scheduled.

Best regards,

Ana

On Tue, Jul 19, 2016 at 3:52 PM, Ian Douglas <ian AT zti.co DOT za> wrote:

hi Ana

On Tuesday 19 July 2016 15:37:05 Ana Emília M. Arruda wrote:

> If you run a differential after a full, then an incremental, but the
> differential one hadn't finished before the incremental one starts, then
> the incremental would check the last full one. So you will have both
> differential and incremental identical.

thanks.

But still not sure that's entirely correct.
I rebuilt a NAS. I had full backup of original.

After rebuilding, paths were slightly different, or at least the file times
were, so when I re-allowed daily incrementals it effectively did a full
backup.

Now, about a month later, it did a differential, again effectively a full
backup.

While differential was running, two daily incrementals were scheduled. I draw
a distinction between 'scheduled' and 'started', I'm not sure if you have the
same difference in your usage.

Anyway it ran two incrementals, both of which backed up the same data (as per
files and filesize in log). Both started hours after the differential was
finished, and the second over an hour after the first finished, with other
jobs inbetween.

My point is that the first incremental was done correctly, the second was an
unnecessary duplication, because (it appears to me), it decided too soon what
needs to be backed up.

Alf said I need to look at the 'duplicate jobs' setting, but that sounds like
a work-around rather than addressing the core issue, ie when does Bacula
decide what needs to be backed up... when scheduled or when run. The second
may be an exact duplicate when scheduled, but the data may have changed by run
time.

Thanks, Ian

--
ian AT zti.co DOT za http://www.zti.co.za
Zero 2 Infinity - The net.works
Phone +27-21-975-7273

------------------------------------------------------------------------------
What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic
patterns at an interface-level. Reveals which users, apps, and protocols are 
consuming the most bandwidth. Provides multi-vendor support for NetFlow, 
J-Flow, sFlow and other flows. Make informed decisions using capacity planning
reports.http://sdm.link/zohodev2dev

_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users