Hi,

I'm using bacula 2.0.3 and I'm having a problem with Bacula recycling 
volumes before I want it to.  This problem has been ongoing since 
version 1.38.something so it's either a very longstanding bug or my 
configuration is broken.

What equipment I have to make sure this all works:

A Dell PowerVault 124T tape autochanger.  Holds 16 tapes, has a barcode 
reader and seems to work perfectly with Bacula.

A scratch pool containing as-yet unpooled tapes.  This scratch pool has 
never had less than 3 tapes in it.  Each day's backup requires 1 tape.


What I want to happen:

Three pools of tapes, Daily, Monthly and Annual.  Daily are recycled 
once a month, Monthly are recycled once a year and Annually are never 
recycled (well, actually they're recycled every 20 years but in the real 
world that means never). 

I don't want to purge any records for tapes which aren't due for 
recycling.  As long as the tape itself hasn't been recycled, I want to 
be able to restore any file from it without resorting to bscan.  I've 
worked out how much space this will require database-wise and I'm 
confident that won't be a problem.

I've had the tape changer less than a year, so in practise I don't 
expect to see any monthly tapes coming up for recycling for a while 
yet.  Instead, I would expect tapes to be taken from the Scratch pool.


What I see happening:

Daily tapes are recycled as necessary at the end of the month.  This is 
fine.
Monthly tapes are being recycled after 2 months, and the Scratch pool is 
ignored.  The backup for 1 April was paused (until I cleared it this 
morning) with the message below:

01-Apr 00:05 gemini-sd: Please mount Volume "000028L3" on Storage Device "LTO3" 
(/dev/nst0) for Job atlas.2007-04-01_00.05.00


Volume 000028L3 was used on 1 February 2007, so shouldn't be coming up 
for recycling until next February. 


How things are currently configured:

My current configuration has the same File and Job retention for every 
job, regardless of which pool they're going to.  The configuration for 
each job is:


  File Retention = 90 days
  Job Retention = 1 month

I'm sure this is wrong, but I'm not sure what it should be.  I want 
every file and job record purged when the volume itself is up for 
recycling and not before.

Pool configurations are:

# Tape pool

Pool {
  Name = DailyTape
  Pool Type = Backup
  Recycle = Yes
  AutoPrune = yes
  Recycle Current Volume = yes
  Volume Retention = 27 days
  Volume Use Duration = 23 hours
}

Pool {
  Name = MonthlyTape
  Pool Type = Backup
  Recycle = Yes
  AutoPrune = Yes
  Volume Retention = 360 days
  Recycle Current Volume = No
  Volume Use Duration = 23 hours
  Storage = Tape
}

Pool {
  Name = AnnualTape
  Pool Type = Backup
  Recycle = No
  AutoPrune = No
  Storage = Tape
  Volume Retention = 20 years
  Recycle Current Volume = No
  Volume use duration = 23 hours
}


The jobs which write to tape are configured thus:

Schedule {
  Name = "WeeklyCycle"
  Run = Level=Full Pool=AnnualTape on 1 jan at 00:05
  Run = Level=Full Pool=MonthlyTape on 1 feb-dec at 00:05
  Run = Level=Full Pool=DailyTape on 2-31 at 00:05
}

Schedule {
  Name = "AmericanWeeklyCycle"
  Run = Level=Full Pool=AnnualTape on 1 jan at 07:05
  Run = Level=Full Pool=MonthlyTape on 1 feb-dec at 07:05
  Run = Level=Full Pool=DailyTape on 2-31 at 07:05
}

Schedule {
  Name = "WeeklyCycleAfterBackup"
  Run = Level=Full Pool=AnnualTape on 1 jan at 07:05
  Run = Level=Full Pool=MonthlyTape on 1 feb-dec at 07:05
  Run = Level=Full Pool=DailyTape on 2-31 at 07:05
}


-- 
U4EA Technologies
http://www.u4eatech.com


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bacula-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-users

Reply via email to