A fileset in bacula does not have to be specific to one job. I understand
that in your case it is specific to one or more jobs, and one server in
particular, but it doesn't have to be job specific. For example you could
back up many systems with one fileset if you only wanted to back up the
same files or directories on the relevant systems (perhaps /home and /root
as a minimum standard). You would have many jobs, one fileset in such a
case.

I am nearly certain that your FD name may be different than originally
specified on your first bacula configuration. From the perspective of
bacula in that case, it has backed up files from someMachine-FD, not
thisMachine-FD.

Maybe try to do a restore in bconsole. See what FD choices are available.
See if one is not your current FD. That would likely be records from your
previously configured FD. You would need to change your new FD's hostname
and also change the bacula configuration files to refer to that hostname.
Case matters for the bacula FD name!

I would also make certain that your NAS mount paths are the same as in your
previous configuration. You can inspect the restore path for some files you
have recorded in your catalog to verify that the exact same path is being
used to mount shares.

Please make absolutely certain that you have carefully exported a backup of
your catalog database! Put this somewhere safe as your first step. Without
this, you will have a much bigger problem to solve. I'm sure you have
already done this, but it's so important that I must say it anyway.

Robert Gerber
402-237-8692
r...@craeon.net

On Fri, Nov 8, 2024, 4:18 PM Samuel Zaslavsky <s...@w4tch.tv> wrote:

> Hi Martin, hi all
>
> Thanks a lot for your answers !
>
> Here's what I can get for one job "BackupARCHIVES" (21 To of datas... the
> small one :)
>
> *estimate level=incremental job=BackupARCHIVES
> Using Catalog "MyCatalog"
> Connecting to Client lto8-fd at localhost:9102
> 2000 OK estimate files=474,483 bytes=21,675,206,135,474
>
>
> Or
>
> *estimate accurate=yes level=incremental job=BackupARCHIVES
> Using Catalog "MyCatalog"
> Connecting to Client lto8-fd at localhost:9102
> 2000 OK estimate files=474,483 bytes=21,675,206,135,474
>
>
> The result should be some Go of datas only...
>
> My fileset :
>
> FileSet {
>   Name = "ARCHIVES"
>   Ignore FileSet Changes = yes
>   Include {
>     Options {
>                 signature = MD5
>                 Accurate = mcso5 # I tried also here with so5
>     }
>         File = /mnt/BIGNAS1/ARCHIVES
>   }
>
>   Exclude {
>         File = "/mnt/BIGNAS1/ARCHIVES/#recycle"
>         File = "/mnt/BIGNAS1/ARCHIVES/.DS_Store"
>   }
> }
>
> And
>
> *select * from fileset where fileset='ARCHIVES' order by CreateTime desc;*
>
> gives one row, the "former" fileset :
>
> "filesetid" "fileset" "md5" "createtime"
>> "44" "ARCHIVES" "q6/YhHgFh/tVI8YO76/asD" "2024-06-07 12:21:00"
>
>
> Also : *SELECT * FROM "public"."job" where name = 'BackupARCHIVES' *
> gives me 19 rows corresponding to the former incremental BackupARCHIVES
> job.
>
> What is weird is that Bacula didn't create another fileset. Maybe it
> doesn't require it for estimate but only for a run ?
>
> What if I run the job, abort it, erase it and give to the new fileset
> (that I suppose would be created ?) the filesetid 44 ? Would it "tell"
> bacula that the former jobs were done with this new fileset (thus
> Incremental would be "repaired" ) ?
> What else could I try ?
>
> Thx a lot !
>
> Sam
>
> Le ven. 8 nov. 2024 à 16:58, Martin Simmons <mar...@lispworks.com> a
> écrit :
>
>> Just to be sure, are you setting level=Incremental in the estimate
>> command?
>>
>> Are there any messages in bconsole after running the estimate command
>> (use the
>> messages command to check)?  In particular, look for "The FileSet ... was
>> modified on ..." or "No prior or suitable Full backup found..."
>>
>> Did you run estimate (or a backup) before adding Ignore FileSet Changes =
>> yes
>> in the FileSet?
>>
>> What is the output of the following sql command?
>>
>> select * from fileset where fileset='...your fileset name...' order by
>> CreateTime desc;
>>
>> __Martin
>>
>>
>> >>>>> On Fri, 8 Nov 2024 14:44:43 +0100, Samuel Zaslavsky said:
>> >
>> > Hello everyone,
>> >
>> > I have a job backing up several hundreds of To.
>> > Job is incremental, every week, so only hundreds of Go, or a few To
>> every
>> > week.
>> > After bad manipulation ( I know, I know...), we lost the bacula
>> > configurations (jobs, fileset, etc.) *but not the database*.
>> >
>> > So we reinstalled bacula, connected it to the old database, recreated
>> Jobs
>> > and filesets etc.
>> > Quite a quick and easy task. In Baculum, everything seems to be back as
>> > before.
>> >
>> > Of course, some fileset configs may have change a little bit but for me
>> it
>> > was not a real issue : I just want the job to backup new files, that's
>> all.
>> > Also I added Ignore FileSet Changes = yes in the FileSet, to be sure it
>> > would not start a new full backup...
>> >
>> > So I was expecting it would "go on" as if nothing happened.
>> > But when I do an estimate, Bacula seems to be wanting to do a brand new
>> > full backup.
>> > This is very annoying, since it would take months, dozens of tapes, etc.
>> >
>> > I would like Bacula to resume at the point where it left off.
>> >
>> > Anyone knows how to achieve that ?
>> >
>> > Thanks a lot for your help !
>> >
>> > Sam
>> >
>> > >
>> >
>>
> _______________________________________________
> Bacula-users mailing list
> Bacula-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
>
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Reply via email to