Re: Issues with Virtual Full backups

Hi,

I think I ran into a similar issue with 15.0.3 - where we do virtual 
fulls for some jobs every X days using the Max Virtual Full Interval = X 
days option.

I noticed that for Fulls where no jobs had run in those X days then it 
would not create a new Full which was just a copy of the previous Full.  
It sounds like that is what you want as well.  In the end, I ended up 
patching the code so it would go ahead and create that new Full even 
with nothing to consolidate since that is how it had worked under our 
earlier (ancient?) 5.2.13 version of Bacula.  We don't use the 
BackupsToKeep feature so it defaults to 0.

In the vbackup.c code, I made it so if BackupsToKeep is 0 then it will 
allow things to continue.

@@ -260,10 +260,14 @@
           Jmsg(jcr, M_INFO, 0, _("Only one Job found in argument. 
Consolidation forced.\n"));

        } else {
-         Jmsg(jcr, M_WARNING, 0, _("BackupsToKeep is set to %u, "
+    /* If BackupsToKeep is not being used then skip this */
+    if (jcr->job->BackupsToKeep > 0)
+    {
+          Jmsg(jcr, M_WARNING, 0, _("BackupsToKeep is set to %u, "
                                     "and/or only %d jobs exits. "
                                     "Consolidation not needed.\n"),
jcr->job->BackupsToKeep, jobids.count);
-         return false;
+          return false;
+    }
        }
     }

In your case, it sounds like you have BackupsToKeep set to 10 and you're 
not getting the above warning/error.  It must be aborting farther 
downstream when it finds those jobs have zero files.  I'm not sure where 
in the code that would be but I could probably find it for you if you 
want.  Not sure if there is a simple patch to avoid the issue or not though.

cheers,

--tom

> I'm running into this issue with Virtual Full backups.  I think I know 
> what the issue is but am not 100% positive and want to bounce this idea...
>
> I have a Full backup, then some number of Incrementals.  Every week I 
> run a Virtual full, rolling in all but the most recent 10 Incremental 
> backups.
> My Virtual Fulls are failing with "No files found to read/Found 0 
> files to consolidate into Virtual Full"
>
> 29-Jul 02:36 bacula-dir JobId 5622: Start Virtual Backup JobId 5622, 
> Job=Taco-Data-A.2025-07-25_21.15.00_33
> 29-Jul 02:36 bacula-dir JobId 5622: Consolidating 
> JobIds=5400,4698,4728,4758,4788,4818
> 29-Jul 02:36 bacula-dir JobId 5622: No files found to read. No 
> bootstrap file written.
> 29-Jul 02:36 bacula-dir JobId 5622: Found 0 files to consolidate into 
> Virtual Full.
> 29-Jul 02:36 bacula-dir JobId 5622: Fatal error: Could not get or 
> create the FileSet record.
>
> What I think is happening is that my disk volume is not being updated 
> (no changes to files, no new files, no files deleted), and when an 
> incremental backup runs it stores nothing on the backup "tape".
>
> 05-Jul 20:19 bacula-dir JobId 4728: Start Backup JobId 4728, 
> Job=Taco-Data-A.2025-06-30_19.10.00_42
> 05-Jul 20:19 bacula-dir JobId 4728: Connected to Storage "FileChanger" 
> at si-scott.miserver.it.umich.edu:9103 
> <http://si-scott.miserver.it.umich.edu:9103/> with TLS
> 05-Jul 20:19 bacula-dir JobId 4728: Using Device "FileChanger-Dev10" 
> to write.
> 05-Jul 20:19 bacula-dir JobId 4728: Connected to Client "taco" at 
> taco.si.umich.edu:9102 <http://taco.si.umich.edu:9102/> with TLS
> 05-Jul 20:19 taco JobId 4728: Connected to Storage at 
> si-scott.miserver.it.umich.edu:9103 
> <http://si-scott.miserver.it.umich.edu:9103/> with TLS
> 05-Jul 20:19 bacula-dir JobId 4728: Sending Accurate information to 
> the FD.
> 05-Jul 20:23 bacula-sd JobId 4728: Elapsed time=00:01:42, Transfer 
> rate=0  Bytes/second
> 05-Jul 20:23 bacula-sd JobId 4728: Sending spooled attrs to the 
> Director. Despooling 0 bytes ...
> 05-Jul 20:23 bacula-dir JobId 4728: Bacula bacula-dir 15.0.3 (25Mar25):
>   Build OS:               x86_64-pc-linux-gnu ubuntu 24.04
> JobId: 4728
>   Job:                    Taco-Data-A.2025-06-30_19.10.00_42
>   Backup Level:           Incremental, since=2025-06-25 09:16:04
>   Client:                 "taco" 15.0.3 (25Mar25) 
> x86_64-pc-linux-gnu,ubuntu,22.04
>   FileSet:                "Taco-Data-A" 2025-05-02 15:31:06
>   Pool:                   "Taco-Incr" (From Job IncPool override)
>   Catalog:                "MyCatalog" (From Client resource)
>   Storage:                "FileChanger" (From Pool resource)
>   Scheduled time:         30-Jun-2025 19:10:00
>   Start time:             05-Jul-2025 20:19:45
>   End time:               05-Jul-2025 20:23:12
>   Elapsed time:           3 mins 27 secs
>   Priority:               10
>   FD Files Written:       0
>   SD Files Written:       0
>   FD Bytes Written:       0 (0 B)
>   SD Bytes Written:       0 (0 B)
>   Rate:                   0.0 KB/s
>   Software Compression:   None
>   Comm Line Compression:  None
>   Snapshot/VSS:           no
>   Encryption:             no
>   Accurate:               yes
>   Volume name(s):
>   Volume Session Id:      619
>   Volume Session Time:    1749606283
>   Last Volume Bytes:      2,295 (2.295 KB)
>   Non-fatal FD errors:    0
>   SD Errors:              0
>   FD termination status:  OK
>   SD termination status:  OK
>   Termination:            Backup OK
>
> So, when it comes time to consolidate it grabs each incremental and 
> finds there's nothing, and the Virtual Full fails.  I can find these 
> "tapes" in my Jobs table, but no joining entries in Media, or JobMedia.
>
> If I scan through the "tape" on the filesystem, I find backups on the 
> tape, only for jobs which actually wrote files.
>
> The end result is that I have a Full backup, which is getting 
> increasingly older, VirtualFulls which fail, so no new Full backup, 
> and once I hit a point where the Full backup hits its prune-date it 
> gets pruned from the file table.  After that, the database knows there 
> was a successful Full, and incrementals still run, but now every file 
> the incremental comes across is "brand new" and my incremental backup 
> is basically a Full backup.
>
>
> -- 
> - Adaptability -- Analytical --- Ideation ---- Input ----- Belief -
> -------------------------------------------------------------------
>          John M. Lockard |  U of Michigan - School of Information
>           Unix Sys Admin |      Suite 205 | 309 Maynard Street
> [email protected] |        Ann Arbor, MI  48104-2211
> www.umich.edu/~jlockard <http://www.umich.edu/%7Ejlockard> |     
> 734-615-8776 | 734-763-9677 FAX
> -------------------------------------------------------------------
> - The University of Michigan will never ask you for your password -
>
>
>
> _______________________________________________
> Bacula-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/bacula-users

_______________________________________________
Bacula-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-users