volume retention and autoprune

[Sunny]

Hi, all

I have a problem with bareos 16.4

I set pools like this

Pool {
Name = pool_4
Pool Type = Backup
Recycle = yes
AutoPrune = yes
Volume Retention = 1 min
Maximum Volume Bytes = 5G
Maximum Volumes = 2
Label Format = pool_4-
}

BUT, When first volume is full and second volume starts to write, first volume's job are removed.

I think I can backup 9G (with 4.5G jobs 2 times).
But when I backup twice, first job gone. (both DB and web-ui)

Please help.

Manual said to me "It keep volume data as long as possible, until no other volume can be written." But, why my volume gone?

When I tested setting like this

Pool {
Name = pool_5
Pool Type = Backup
Recycle = no
AutoPrune = no
Volume Retention = 1 hour
Maximum Volume Bytes = 5G
Maximum Volumes = 2
Label Format = pool_5-
}

There is no such a problem. But I want to Recycle the volume automatically.
Please Help. Thanks.

[Sunny]

When I run bareos on debug mode, I got this.
Why bareos purge my job?? Although I have appendable volume?

bareos-dir (100): autoprune.c:89-1130 Prune volumes PoolId=172
bareos-dir (100): autoprune.c:125-1130 Scratch pool(s)=
bareos-dir (100): autoprune.c:150-1130 query=SELECT DISTINCT MediaId,LastWritten FROM Media WHERE (PoolId=172 OR RecyclePoolId IN (172)) AND MediaType='File' ORDER BY LastWritten ASC,MediaId
bareos-dir (100): autoprune.c:156-1130 Volume prune num_ids=1
bareos-dir (100): autoprune.c:164-1130 Get record MediaId=174
bareos-dir (100): autoprune.c:169-1130 Examine vol=bareos3_linux2-0174
bareos-dir (100): autoprune.c:178-1130 Add prune list MediaId=174 Volume bareos3_linux2-0174
bareos-dir (40): ua_prune.c:889-1130 Query=SELECT DISTINCT JobMedia.JobId FROM JobMedia,Job WHERE MediaId=174 AND Job.JobId=JobMedia.JobId AND Job.JobTDate<1514510550
bareos-dir (40): ua_prune.c:932-1130 accept JobId[0]=1129
bareos-dir (100): autoprune.c:180-1130 Num pruned = 1
bareos-dir (150): ua_purge.c:361-1130 num_ids=1
bareos-dir (150): ua_purge.c:374-1130 Add id=1129
bareos-dir (150): ua_purge.c:377-1130 num_ids=0
bareos-dir (40): ua_purge.c:334-1130 Delete File sql=DELETE FROM File WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:338-1130 Delete BaseFiles sql=DELETE FROM BaseFiles WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:348-1130 Mark purged sql=UPDATE Job SET PurgedFiles=1 WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:495-1130 Delete JobMedia sql=DELETE FROM JobMedia WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:499-1130 Delete Log sql=DELETE FROM Log WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:503-1130 Delete RestoreObject sql=DELETE FROM RestoreObject WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:507-1130 Delete PathVisibility sql=DELETE FROM PathVisibility WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:511-1130 Delete NDMPJobEnvironment sql=DELETE FROM NDMPJobEnvironment WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:515-1130 Delete JobStats sql=DELETE FROM JobStats WHERE JobId IN (1129)
bareos-dir (40): ua_purge.c:469-1130 Upgrade copies Log sql=CREATE TEMPORARY TABLE cpy_tmp AS SELECT MIN(JobId) AS JobId FROM Job WHERE Type='C' AND ( PriorJobId IN (1129) OR PriorJobId IN ( SELECT PriorJobId FROM Job WHERE JobId IN (1129) AND Type='B' ) ) GROUP BY PriorJobId
bareos-dir (40): ua_purge.c:523-1130 Delete Job sql=DELETE FROM Job WHERE JobId IN (1129)

I'm stuck of this. please answer me.

[Dan]

Sunny -

The problem that you are having is with your Volume Retention settings. Your debug output doesn't show that your job is being 'purged', it shows that it is being 'pruned'. With "AutoPrune=yes" and "VolumeRetention=1 min", Bareos will prune every job that is more than 1 minute old. If the volume is then empty it will mark it purged and will recycle it. Pick a time in the future at which point the backup data is no longer important to you and set that as the Volume Retention (i.e. 180 day, 1 year). That will resolve your issue.

[Sunny]

Thanks. Dan.

I Think volume will pruned when there is no more space in pool.

I set volume retention long period, and the problem solved.

Thank you for answering!