Am 27-Sep-2021 11:46:31 +0200 schrieb s...@w4tch.tv:
>
> Hello everyone,
>
> Anyone could help me here ?
>
> Thanks a lot !
>
> Samuel
>
> Le jeu. 23 sept. 2021 à 12:46, Samuel Zaslavsky <s...@w4tch.tv> a écrit :
> >
> > Hello everyone,
> > We have set up a system with a tape library and Bacula to back up / archive
> > some NASs. The idea is as simple as possible: no files expire, I just have
> > to make sure that in the event of a disaster on a NAS, we will be able to
> > recover all the files.
> > (So I have a pretty basic knowledge of Bacula, enough to set up this very
> > simple project, but didn't need to become an expert ... my apologies in
> > advance if I did mistakes or if I talk nonsense :))
> >
> > So I have "Incremental" jobs that run every night.
> > During the summer (and the holidays ...) the tape library found itself
> > short of LTO, and for a period of about 15 days, jobs were created and put
> > on hold ("Created but not yet running ".)
> >
> > Naively, I told myself that making LTOs available would solve the problem.
> > In a sense, this is the case, because jobs have indeed started up again.
> >
> > But there is a big but !! It seems that the jobs have rewritten several
> > times (as many times as jobs not launched during the period) almost the
> > same data ...
> > So to my surprise, I still find myself running out of LTO a few days after
> > settling the pb.
> >
> > 1 / So I realize that I should have canceled all the "pending" jobs apart
> > from the last one created, is that right? Or if not, how to handle this
> > case? (this is important because I am in this case again!)
> >
> > 2 / How to recover the wasted space (almost 10 LTO8 !!)
> > Specifically, how do you go about identifying exactly which volumes / jobid
> > are to be "deactivated" and how do you do that? How to end with a clean
> > incremental job running smoothly ?
> >
> > Thank you very much in advance for your help!
> >
> > Samuel
> >
Hello Samuel,
I'm a little bit out of time by now, but I will give my best to help you to
refine the questions a little bit.
What do you mean by "no files expire?"
The files on the NASs or the volume-, file-, job- retention of bacula?
Normally when I start an incremental job and nothing changed on the NAS the
written Bytes of the incremental-job are zero(status dir on the bconsole). Is
this correct for you also? Are you really doing incremental backups?
With that in mind the second idea I have is that bacula creates new volumes on
the tape(s) since the volume-use duration is set to short.
Documentation states:
"You might use this directive, for example, if you have a Volume used for
Incremental
backups, and Volumes used for Weekly Full backups. Once the Full backup is
done, you
will want to use a different Incremental Volume. This can be accomplished by
setting the
Volume Use Duration for the Incremental Volume to six days. I.e. it will be
used for the
6 days following a Full save, then a different Incremental volume will be used.
Be careful
about setting the duration to short periods such as 23 hours, or you might
experience
problems of Bacula waiting for a tape over the weekend only to complete the
backups
Monday morning when an operator mounts a new tape."
Chapter: "Configuring the Director" under "Volume Use duration" or Page: 274f
I'm not an expert, PLEASE take it with a grain of salt
1. Being honest I don't know. But if the jobs didn't run and do *exactly* the
same I would go ahead and delete them except one.
2. 10 LTO8 divided by 15 is quiet much data for a daily incremental backup.
In the bacula-dir.conf you specified the director ressource type "Messages"
there is a option called "append"
A part of my bacula-dir.conf:
# WARNING! the following will create a file that you must cycle from
# time to time as it will grow indefinitely. However, it will
# also keep all your messages if they scroll off the console.
append = "/var/log/bacula/bacula.log" = all, !skipped
console = all, !skipped
At the end "all, !skipped" are the types or classes of messages which go into
it. They are described in more detail in the "Messages Resource"-Chapter:
https://www.bacula.org/11.0.x-manuals/en/main/Messages_Resource.html
If I type the "messages"-command in the bconsole the output is in my case in
both cases the same.
Furthermore:
For every job there is a list. Under "Volume name(s):" are the volumes listed
that bacula used in that job.
Notice in my case with a 0Byte incremental job this list empty -> I'm not sure,
but bacula shouldn't create empty volumes.
Bacula states if you run "delete" in the bconsole:
"In general it is not a good idea to delete either a
Pool or a Volume since they may contain data."
My uneducated guess where things could went wrong and I would go ahead to check
condensed:
-are you really running incremental backups?
-Volume Use Duration is set quite short?
-maybe a problem with different pools?
-are the jobs set properly?
-multiple problems?
Back to your second question:
I wouldn't delete them and set the volume-,file-,job- retention properly and
let bacula sort it out.
I hope that helps a little bit. If not, please wait a little bit longer and you
may get an email from a person with more knowledge than me.
I would appreciate it, if someone with more knowledge scans through my email
and states things out, which may be wrong.Thanks!
Kind regards
Sebastian
-------------------------------------------------------------------------------------------------
FreeMail powered by mail.de - MEHR SICHERHEIT, SERIOSITÄT UND KOMFORT
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
