suddenly problems with full backups (bareos 24.05)

[Markus Dubois]

..since recently i have the issue that a long running full backup (15 TB) runs ....

...until the daily schedule starts the expected incremental job.
earlier in the time when everything was working, the incremental job wanted to start a full, as it doesn´t find one, but got blocked by the "no duplicates" directive.

this no longer works now.

I have the situation, that the full job got silently killed and the new earlier incremental, "now wanna be full" job takes over.
This is unfortunate as until now i had no "Max Run" time configured, what should mean "forever". Now i´ve set 48h and look tonight if the full runs trough.

Here is my job config and my jobdefs

Job {
  Name = "AIbackup-omvserver"
  Client = "omvserver"
  FileSet = omvserver
  Type = Backup
  Level = Incremental
  Schedule = "AISchedule"
  Storage = FileCons
  Priority = 50
  Messages = Standard
  Pool = AI-Incremental
  Max Run Time = 48h
  Spool Attributes = yes
  Maximum Concurrent Jobs = 1
  Full Backup Pool = AI-Consolidated
  Incremental Backup Pool = AI-Incremental
  Accurate = yes
  Allow Mixed Priority = no
  Allow Duplicate Jobs = no
  Always Incremental = yes
  Always Incremental Job Retention = 7 days
  Always Incremental Keep Number = 7
  Always Incremental Max Full Age = 11 days
  Max Virtual Full Interval = 14 days
  Run Script {
    Console = ".bvfs_update jobid=%i"
    RunsWhen = After
    RunsOnClient = No
  }
Run Script {
    Command = "/var/lib/bareos/scripts/jobcheck.sh"
    RunsWhen = Before
    RunsOnClient = No
    Fail job on error = No
  }
  Run Script {
    Command = "rm -f /var/lib/bareos/scripts/job.running"
    RunsWhen = After
    RunsOnClient = No
    Fail job on error = No
  }
}

JobDefs {
  Name = "DefaultJob"
  Type = Backup
  Level = Incremental
  Client = nasbackup
  FileSet = "Catalog"                     # selftest fileset                            (#13)
  Schedule = "WeeklyCycle"
  Storage = File
  Messages = Standard
  Pool = AI-Incremental-Catalog
  Priority = 10
  Write Bootstrap = "/var/lib/bareos/%c.bsr"
  Run Script {
    Console = ".bvfs_update jobid=%i"
    RunsWhen = After
    RunsOnClient = No
  }
  Full Backup Pool = AI-Consolidated-Catalog                  # write Full Backups into "Full" Pool         (#05)
}

[Markus Dubois]

no, also adding the Max Run Time = 48h didn´t help. again it´s 19:00 and the incremental job gets upgraded to full and overtakes everything and the running full stops without error.
this worked for years. the only thing i did was the allignment of the pools, so that everything goes into AI-Consolidated (see above), so that virtual fulls are working.

Now physical fulls have those issues....

[Bruno Friedmann (bruno-at-bareos)]

Hello Markus,

From what I can see in your configuration, it look like you missed to setup one off the cancel instruction

see https://docs.bareos.org/master/_images/duplicate-real.svg

I personally have my jobs setup with 

  Allow Duplicate Jobs = no

  Cancel Lower Level Duplicates = yes
  Cancel Queued Duplicates = yes

Cancel Running Duplicates = no ( not written as it is the default )

and didn't seen what you experiment (it doesn't mean it can't exist) but we would have to find then the why.

[Markus Dubois]

thanks, i´ll read and implement this accordingly