Kevin Keane a écrit :
> Yann Cézard wrote:
>
>> Kevin Keane a écrit :
>>
>>> If my math is right, the 12th is the 13th day of the retention
>>> period. Since the previous job finished at 22:43, and your new job
>>> started at 22:38, the rentention period hadn't elapsed yet, by about
>>> five minutes. The reason it sometimes works and sometimes doesn't is
>>> simply the timing between the jobs.
>>>
>> Actually, I don't agree with your maths, sorry :)
>> The 12th was the 14th day of the retention period.
>> 29th 22:43:13 => start of retention, day/hour 0
>> 30th 22:43:13 => 1st day
>> 31th 22:43:13 => 2nd day
>> 1st 22:43:13 => 3rd day
>> 2nd 22:43:13 => 4th day
>> ...
>> 11th 22:43:13 => 13th day : job could be pruned, and then the volume
>> 12th 22:38 => job is still there, volume is still not recycled ???
>>
>> So basically it looks more like a Job pruning problem ?
>>
> Oops. I think you are right. What are the volume, job and file retention
> times? I didn't see those two values in the tables you sent earlier;
> maybe I missed it. The only retention time I saw was the one for the
> pool - which really has no direct effect at all here. Maybe one of the
> files still had an unexpired retention time? In that case, the job won't
> get purged even if the job itself has passed its retention time.
>
Actually, File Retention is configured 13 days for every clients.
But you know what ? that's a very good point ! I explain myself :
At the start of my migration, I had 2 director crashes, caused by old
configuration
files for Messages which were mentioning /var/bacula/log instead of
/var/lib/bacula/log
I noticed the volumes that were used at the time of the crash had a
LastWritten field
value of "0000-00-00 00:00:00" which was causing trouble recycling, I
already fixed
that days ago, but I didn't thought to look at the File table !
I'll have a look at that, and keep you informed, thanks for pointing it
up to me !
> Wow. That's pretty big. I didn't check in your original post, but is it
> possible that some of the full backups take more than 24 hours? That
> might throw off the schedule, too
Sure it is big !
In fact, all data is backupped from about 60 differents clients.
The biggest job, which could not be done in 24 hours, is splitted in 7 jobs
backing up a part of the total file system, each Full running on a
different day.
And with data volume keeps growing on it, I will probably have to move on
a greater basis schedule, but that is another story :)
Thanks again,
Yann
--
Yann Cézard - Administrateur Systèmes Serveurs
Centre de Ressources Informatiques - http://cri.univ-pau.fr
Université de Pau et des Pays de l'Adour - http://www.univ-pau.fr
------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users
|
