Re: [Bacula-users] Migrate/copy from one pool to different pools based on schedule

[Hugo L. <hug...@gm...>]

2012/6/28 Mario Moder <li...@de...>:
> Am 28.06.2012 13:31, schrieb Hugo Letemplier:
>> 2012/6/28 Mario Moder <li...@de...>:
>>> Hi Bacula Community.
>>>
>>> Simple setup: We have one "file-pool" which is used on a weekly
>>> schedule (saturdays). We have another "tape-pool" and have
>>> configured a
>>> copy job, scheduled every monday, which copies all uncopied jobs
>>> from
>>> the "file-pool" to the "tape-pool". This works, no problems so far.
>>>
>>> Now I want to change the copy schedule so that it copies the
>>> "file-pool" jobs to a "monthly-tape-pool" every 1st monday and
>>> copies
>>> the jobs from "file-pool" to a "weekly-tape-pool" every 2nd to 5th
>>> monday (for some kind of "father-son" tape rotation)
>>>
>>> As far as I understand from the docs there may only be one "Next
>>> pool"
>>> directive in the "file-pool". For the above to work I would somehow
>>> need
>>> to overide the "Next pool" from within the schedule configuration
>>> (use
>>> "Next Pool = monthly-tape-pool" on 1st monday and use "Next pool =
>>> weekly-tape-pool" on the remaining mondays. Is this possible?
>>>
>
>>
>> You should use a "dummy" pool that you will use to run your copy job
>> and this dummy pool's next pool should be set to the destination of
>> your copy job
>>
>> I hope you will understand
>>
>
> Thanks for your answer, Hugo.
>
> How does the copy job know then, _from_ which pool it should copy, when
> the "dummy" pool has no volumes?
>
> I should have included my config for clarity, here it is:
>
> # bacula-dir.conf
> # File pool
> Pool {
>   Name = File
>   Pool Type = Backup
>   Recycle = yes                       # Bacula can automatically
> recycle Volumes
>   AutoPrune = yes                     # Prune expired volumes
>   Volume Retention = 30 days
>   Maximum Volume Bytes = 100G          # Limit Volume size to something
> reasonable
>   Maximum Volumes = 10               # Limit number of Volumes in Pool
>   Action On Purge = Truncate
>   Next Pool = tape-weekly-pool
>   Storage = file-stor
> }
>
> # tape weekly pool
> Pool {
>   Name = tape-weekly-pool
>   Pool Type = Backup
>   Recycle = yes
>   AutoPrune = yes
>   Volume Retention = 25 days
>   Volume Use Duration = 3 days
>   Recycle Current Volume = yes
>   LabelFormat = "tape-weekly-"
>   Maximum Volumes = 4
>   Storage = tape-stor
> }
>
> # tape monthly pool
> Pool {
>   Name = tape-monthly-pool
>   Pool Type = Backup
>   Recycle = yes
>   AutoPrune = yes
>   Volume Retention = 5 months
>   Volume Use Duration = 3 days
>   Recycle Current Volume = yes
>   LabelFormat = "tape-monthly-"
>   Maximum Volumes = 6
> }
>
> Job {
>   Name = "copy-job"
>   Type = Copy
>   Level = Full
>   Client = bacula-fd
>   FileSet = "Full Set"
>   Messages = Standard
> #  Schedule = copy-schedule
>   Pool = File
>   Selection Type = PoolUncopiedJobs
> }
>
>
>
> The "fictional" schedule (it doesn't work this way, but it should be
> easier to understand) should look like this:
>
> Schedule {
>   Name = "copy-schedule"
> #  Run = Level=Full Pool=File NextPool=tape-monthly-pool 1st mon at
> 10:00
> #  Run = Level=Full Pool=File NextPool=tape-weekly-pool 2nd-5th mon at
> 10:00
> }
>
> I don't quite understand how the dummy pool you suggested should look
> like, it would be nice if you could elaborate on that.
>
> Thanks for your answer
>
> Mario

Ah ok, I use a selection in Postgresql database to get the jobid list
for my copy job

Hugo