Backup Central

Hi the list

I have got a big space of 1,6 TB that contain all my backups jobs
I have often a filesystem full that is caused by two behaviors of Bacula:
- If a job run in a pool, it wont prune jobs in other pools than the one of the running job, so it cant win some space from volumes or jobs that can be pruned in the other pools although all theses jobs are physically stored in the same file system
- If the file system is full, bacula will create volume, trying to write it, fail, labeling another volumes until it reaches the max possible volume in the pool.
- In a same way, bacula waits for the volume to be mounted but this will never work because file system is full or volumes are all with a error status.

To force free space recovery I use a script that runs everydays: it scans my pools, getting a volume list, pruning all theses volumes, truncating them if necessary.
However, sometimes I have a big job (with level full) accumulation and so it causes a big kaboom all jobs failing, changing volumes to error status, reaching the max volumes/pool limit that then cause next jobs to fail…until an operator action to purge all volumes that got error status, and then running the maintenance script to recover some space…

Is it possible to check a free space ratio on the storage daemon, fail a job if required instead of failing also the volumes, waiting for a mount… and thus avoid a lot of successive error that could cause miscomprehension for the operator. Else have you got a solution to avoid my issue. I will have more space in few weeks but my backups are also being much larger due to enterprise expansion.

Can you help me

Thanks

Hugo
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Bacula-users mailing list
Bacula-users < at > lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

On Thu, Jun 14, 2012 at 11:25:03PM +0200, Yougo wrote:

Is it possible to check a free space ratio on the storage daemon,
fail a job if required instead of failing also the volumes, waiting
for a mount… and thus avoid a lot of successive error that could
cause miscomprehension for the operator. Else have you got a
solution to avoid my issue. I will have more space in few weeks but
my backups are also being much larger due to enterprise expansion.

For the moment I think your best bet would be a RunBeforeJob script
that checks the free space on the filesystem, retrieves the size of
the last full backup for the client being backed up and then does some
comparisons based on these parameters.

If the question is allowed: Why do you have so many pools defined? I
usually only have my incremental an full pools, but of course once the
PHBs start bugging you with different backup "classes" and
"strategies", that's the first thing that usually goes out of the
window.

All the best, Uwe



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Bacula-users mailing list
Bacula-users < at > lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

2012/6/15 Uwe Schuerkamp <uwe.schuerkamp < at > nionex.net>:

On Thu, Jun 14, 2012 at 11:25:03PM +0200, Yougo wrote:

Is it possible to check a free space ratio on the storage daemon,
 fail a job if required instead of failing also the volumes, waiting
 for a mount… and thus avoid a lot of successive error that could
 cause miscomprehension for the operator. Else have you got a
 solution to avoid my issue. I will have more space in few weeks but
 my backups are also being much larger due to enterprise expansion.

For the moment I think your best bet would be a RunBeforeJob script
that checks the free space on the filesystem, retrieves the size of
the last full backup for the client being backed up and then does some
comparisons based on these parameters.

If the question is allowed: Why do you have so many pools defined? I
usually only have my incremental an full pools, but of course once the
PHBs start bugging you with different backup "classes" and
"strategies", that's the first thing that usually goes out of the
window.

All the best, Uwe


I have different classes of jobs because I have about 4 different
kind of backup
I split data from the system backup ( I need b, so we had 6 pools also
I have an old mail system based on big database files that need to be
kept for a shorter time

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Bacula-users mailing list
Bacula-users < at > lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

2012/6/15 Hugo Letemplier <hugo.let.35 < at > gmail.com>:
2012/6/15 Uwe Schuerkamp <uwe.schuerkamp < at > nionex.net>:

On Thu, Jun 14, 2012 at 11:25:03PM +0200, Yougo wrote:

Is it possible to check a free space ratio on the storage daemon,
 fail a job if required instead of failing also the volumes, waiting
 for a mount… and thus avoid a lot of successive error that could
 cause miscomprehension for the operator. Else have you got a
 solution to avoid my issue. I will have more space in few weeks but
 my backups are also being much larger due to enterprise expansion.

For the moment I think your best bet would be a RunBeforeJob script
that checks the free space on the filesystem, retrieves the size of
the last full backup for the client being backed up and then does some
comparisons based on these parameters.

If the question is allowed: Why do you have so many pools defined? I
usually only have my incremental an full pools, but of course once the
PHBs start bugging you with different backup "classes" and
"strategies", that's the first thing that usually goes out of the
window.

All the best, Uwe





I have different classes of jobs because I have about 4 different
kind of backup
I split data from the system backup ( I need b, so we had 6 pools also
I have an old mail system based on big database files that need to be
kept for a shorter time

Sorry my mail was sent before I finished writing

Also I have got different kind of content that should be kept for a
very long time.
To finish I have an Archive pool that stores job copies on LTO Tapes

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Bacula-users mailing list
Bacula-users < at > lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users