So it seems from the source code that the selection depends on a new
job being created for the copy, which references the original via the
PriorJobId self join. This doesn't lend itself to easy faking out with
a SQL update, it seems to me.
So, what can I do?
Run the copy jobs to storage director /dev/null ?
I'd like to find a solution to this.
On 4 April 2014 09:19, Gary Cowell <gary.cowell AT gmail DOT com> wrote:
> Hello
>
> I have bacula backing up to networked storage.
>
> For offsite storage once a week I run a PoolUncopiedJobs job to copy
> the uncopied FULL pool jobs to USB drives for going offsite, job is :
>
> JobDefs {
> Name = CopyDiskToUSB
> Type = Copy
> Messages = Standard
> Client = None
> FileSet = None
> Selection Type = PoolUncopiedJobs
> Maximum Concurrent Jobs = 10
> #Storage = File
> SpoolData = No
> Allow Duplicate Jobs = Yes
> Cancel Queued Duplicates = No
> Cancel Running Duplicates = No
> Priority = 13
> }
> Job {
> Name = CopyDiskToUSBFullBackups
> Enabled = Yes
> Schedule = MondayMorning
> Pool = FullPool
> JobDefs = CopyDiskToUSB
> Priority = 6 # complete these before incrementals
> Maximum Concurrent Jobs = 4
> }
>
> Pool {
> Name = FullPool
> Pool Type = Backup
> Recycle = yes
> Recycle Pool = FullPool
> AutoPrune = yes
> Storage = FullFileStorage
> Volume Retention = 3 weeks
> Maximum Volume Bytes = 100G
> Maximum Volumes = 40
> LabelFormat = "FULL_"
> Next Pool = Offsite
> Action On Purge = Truncate
> }
>
>
> And this works, the FULL pool jobs get copied to the Offsite pool,
>
> BUT ...
>
> There are times, for one reason or another, perhaps the USB enclosure
> has a problem, or some other reason, doesn't matter what it is, that
> the uncopied jobs don't copy.
>
> This creates a problem because next time the FULL backups happen, we
> end up with /even more/ uncopied jobs to copy, then there isn't enough
> time, or USB storage to do them all, but we don't really want the
> 'stale' uncopied jobs, we only need the latest FULL jobs.
>
> So can we either
>
> A) only select the Most Recent FULL backup for copying with this method or
>
> B) run something (sql?) to mark all current Uncopied jobs as COPIED so
> they don't get selected next time?
>
> Either one will do. Option A would be better, as we'd know which
> weren't copied, but practically speaking, either works.
>
> Or, something else to achieve the same?
>
> Thanks
------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/NeoTech
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users
|
