Re: [Bacula-users] volumes, pools, media types and labels

2024-06-21 Thread Stefan G. Weichinger

On 21.06.24 16:31, Radosław Korzeniewski wrote:

There are multiple places it could be wrong. You will never know unless 
you will get a whole picture.

Now we only have a few puzzle pieces which do not match together.


Sure, I will compile config and logs asap and come back here.
thanks



___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] volumes, pools, media types and labels

2024-06-21 Thread Radosław Korzeniewski
Hello,

czw., 20 cze 2024 o 17:46 Stefan G. Weichinger  napisał(a):

> Am 20.06.24 um 17:32 schrieb Radosław Korzeniewski:
>
> > Great! So what job logs say about a volume selection when it selects "a
> > wrong volume"?
>
> Well, it looks for a volume named "Vol-xxx" in the tape loader, which is
> wrong.
>

Any chance you can share your job log here?


>
> > Your configuration seems to be fine, but what Bacula says about your
> > configuration - check bconsole command show job=...
> > Did you reload your configuration after any change?
>
> I restarted the daemons now and then, if that's the same.
>
> > Are your volumes labeled correctly?
>
> I think so. That one "ghost volume" is created by Bacula somehow.
>

Any chance you can share your tape volume list here?


>
> > Pool "daily" should not create a "Vol-" volume in the "File" storage
> > ...
> > it shouldn't use "File" at all.
> >
> >
> > Yes, if you configured it as described.
> > I never had such a mistake in Bacula when correctly configured, but I've
> > seen this kind of mistake when incorrectly configured.
>
> Well, that's why I posted the config lines ;-)
> I agree and know that something is incorrectly configured. The question
> is: what?
>

There are multiple places it could be wrong. You will never know unless you
will get a whole picture.
Now we only have a few puzzle pieces which do not match together.

best regards
-- 
Radosław Korzeniewski
rados...@korzeniewski.net
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Limit job rescheduling

2024-06-21 Thread Lionel PLASSE
Hello,
I read your post  :

>
>   > i set 'Max Run Sched Time = 8 hours' to prevent conflicts).
>
>   That's something I had seen and pondered.
>   However, it will stop a job that is already running, won't it?
>   So, in my case, if a job started late in the afternoon it wouldn't be 
> allowed to complete. Right?

And maybe you can study those options behavior and adjust in your case:
 Duplicates and priorities
AllowMixedPriority:
AllowDuplicateJobs:
CancelLowerLevelDuplicates:
CancelQueuedDuplicates:
CancelRunningDuplicates:





-Message d'origine-
De : Andrea Venturoli 
Envoyé : vendredi 21 juin 2024 10:42
À : bacula-users@lists.sourceforge.net
Objet : Re: [Bacula-users] Limit job rescheduling

On 6/20/24 17:05, Marco Gaiarin wrote:

> Ciao Andrea.

Ciao.



> i set 'Max Run Sched Time = 8 hours' to prevent conflicts).

That's something I had seen and pondered.
However, it will stop a job that is already running, won't it?
So, in my case, if a job started late in the afternoon it wouldn't be allowed 
to complete. Right?



> So i've moved most client on static ip.

I always use static DHCP entries for clients I know about :)



  bye & Thanks
av.


___
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


Re: [Bacula-users] Again LTO9 and performances...

2024-06-21 Thread Josh Fisher via Bacula-users




On 6/20/24 18:58, Bill Arlofski via Bacula-users wrote:

On 6/20/24 8:58 AM, Marco Gaiarin wrote:

Once that is hit, the spoofles are written to tape, during which active
jobs have to wait because the spool is full.

There's no way to 'violate' this behaviour, right?! A single SD process
cannot spool and despool at the same time?


An SD can be spooling multiple jobs wile *one* and only one Job spool 
file is despooling to one drive.


Add another drive and and the same is still true, but the SD can now 
be despooling two jobs at the same time while other jobs are spooling, 
and so on as you add drives.




Except when the MaximumSpoolSize for the Device resource is reached or 
the Spool Directory becomes full. When there is no more storage space 
for data spool files, all jobs writing to that device are paused and the 
spool files for each of those jobs are written to the device, one at a 
time. I'm not sure if every job's spool file is written to tape at that 
time, or if the forced despooling stops once some threshold for 
sufficient space has been freed up. So, more accurately, other jobs 
continue spooling IFF there is sufficient space in the spool directory.




___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Verify Jobs not pruning cache when using S3 storage

2024-06-21 Thread Martin Reissner

Hello Eric,

yes, this actually helps a lot. I was aware of the "cloud prune" command and 
was already considering scripting something with that. What
I wasn't aware of is the "allpools" keyword which really makes things easier, 
so thank you for bringing it to my attention!

All the best,

Martin

On 19.06.24 19:01, Eric Bollengier wrote:

Hello Martin,

On 6/19/24 12:57, Martin Reissner wrote:

Hello,

we're running a Bacula 15.0.2 setup which stores everything in S3 storage and 
due to having a lot of data to backup every day
we use

   Truncate Cache = AfterUpload
   Upload = EachPart

in our "Cloud" ressources, to ensure the systems running the SDs do not run out 
of diskspace. This works pretty well but I recently
started to configure Verifyjobs with

   Level = VolumeToCatalog

for some important jobs and it looks like when running the Verifyjob, data gets 
transferred from S3 to the directory specified
as "Archive Device" in the "Device" ressource, sometimes also known as 
"cloudcache" and is then compared to the respective data in the catalog.
So far so good, but it seems that even after a successful verification the 
volume data remains there, probably until the volume is reused and for
large jobs this takes up quite some space which causes diskspace issues on the 
SD.

Am I maybe missing something there and is there some option that can be used to 
prune the cache after a (successful) verifyjob?


Yes, you have a cloud command to prune the cache, I suggest to run it at a
regular interval inside an admin job for example. (cloud prune allpools 
storage=XXX)

Hope it helps!

Best Regards,
Eric



--
Wavecon GmbH
 
Anschrift:  Thomas-Mann-Straße 16-20, 90471 Nürnberg

Website:www.wavecon.de
Support:supp...@wavecon.de
 
Telefon:+49 (0)911-1206581 (werktags von 9 - 17 Uhr)

Hotline 24/7:   0800-WAVECON
Fax:+49 (0)911-2129233
 
Registernummer: HBR Nürnberg 41590

GF: Cemil Degirmenci
UstID:  DE251398082



___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Limit job rescheduling

2024-06-21 Thread Andrea Venturoli

On 6/20/24 17:05, Marco Gaiarin wrote:


Ciao Andrea.


Ciao.




i set 'Max Run Sched Time = 8 hours' to prevent conflicts).


That's something I had seen and pondered.
However, it will stop a job that is already running, won't it?
So, in my case, if a job started late in the afternoon it wouldn't be 
allowed to complete. Right?





So i've moved most client on static ip.


I always use static DHCP entries for clients I know about :)



 bye & Thanks
av.


___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] LTO-8 tape report error

2024-06-21 Thread Radosław Korzeniewski
Hello,

czw., 20 cze 2024 o 12:56 Adam Weremczuk 
napisał(a):

> OMG...
>
> The only reason the script started failing was a different device
> assignment caused by a reboot:
>
> ls -al /dev | grep sg.
> crw-rw  1 root tape 21,   0 Jun 19 12:25 sg0
> crw-rw  1 root disk 21,   1 Jun 12 17:42 sg1
>
>
It is best to use persistent device names instead of above. You should ask
udev (with udevadm info --query=all --name /dev/sg0) about it.
This device path never changes.

best regards
-- 
Radosław Korzeniewski
rados...@korzeniewski.net
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users