Bacula-users

Re: [Bacula-users] A question about Virtual Fulls

2011-03-17 17:12:18
Subject: Re: [Bacula-users] A question about Virtual Fulls
From: Mike Carlson <carlson39 AT llnl DOT gov>
To: Silver Salonen <silver AT ultrasoft DOT ee>
Date: Thu, 17 Mar 2011 14:09:28 -0700
On 03/15/2011 10:05 PM, Silver Salonen wrote:
> On Wednesday 16 March 2011 01:32:29 Mike Carlson wrote:
>> I've been semi-successful with doing daily incrementals, and then a
>> VirtualFull at the end of the month. However, after  two months of doing
>> this, the virtual full wants lasts months virtual full. This isn't
>> working because the File device cannot read and write simultaneously to
>> itself... It appears to be a chicken and the egg problem that I don't
>> know how to work around.
>>
>> Here is my schedule:
>>
>>      Schedule {
>>           Name = "1am"
>>           Run = Incremental mon-sun at 01:30
>>      }
>>
>>      Schedule {
>>           Name = "1am-vf"
>>           Run = VirtualFull 3rd sun at 01:30
>>      }
>>
>> A client has two jobs, one for the first Full and incrementals, and
>> another job for the virtual full.
>>
>>      Job {
>>           Name = "client-a"
>>           Type = Backup
>>           Level = Incremental
>>           FileSet = "client-a FileSet"
>>           Client = "client-a"
>>           Storage = FileStorageD2
>>           Pool = client-a-File
>>           Schedule = "1am"
>>           Messages = Standard
>>           Priority = 10
>>           Write Bootstrap = "/var/db/bacula/%c.bsr"
>>           Maximum Concurrent Jobs = 10
>>           Reschedule On Error = yes
>>           Reschedule Interval = 1 hour
>>           Reschedule Times = 8
>>           Allow Duplicate Jobs = no
>>      }
>>
>>      Job {
>>           Name = "client-a-vf"
>>           Type = Backup
>>           Level = VirtualFull
>>           FileSet = "client-a FileSet"
>>           Client = "client-a"
>>           Storage = VF_FileStorageD2
>>           Pool = client-a-File
>>           Schedule = "1am-vf"
>>           Messages = Standard
>>           Priority = 10
>>           Write Bootstrap = "/var/db/bacula/%c.bsr"
>>           Maximum Concurrent Jobs = 10
>>           Reschedule On Error = yes
>>           Reschedule Interval = 1 hour
>>           Reschedule Times = 8
>>           Allow Duplicate Jobs = no
>>      }
>>
>>
>> Each client has two pools and two devices it can write to:
>>
>>      Pool {
>>           Name = client-a-File
>>           Pool Type = Backup
>>           Recycle = yes
>>           AutoPrune = yes
>>           Volume Retention = 1 months
>>           Maximum Volume Bytes = 5G
>>           Maximum Volumes = 100
>>           LabelFormat = "client-aFileVol"
>>           Maximum Volume Jobs = 5
>>           Next Pool = client-a-vf
>>           Storage = FileStorageD2
>>      }
>>
>>      Pool {
>>           Name = client-a-vf
>>           Pool Type = Backup
>>           Recycle = yes
>>           AutoPrune = yes
>>           Volume Retention = 1 months
>>           Maximum Volume Bytes = 5G
>>           Maximum Volumes = 100
>>           LabelFormat = "client-aVFVol"
>>           Maximum Volume Jobs = 5
>>           Storage = VF_FileStorageD2
>>      }
>>
>> One the filesystem, it ends up looking like this (I have a drive##, and
>> a separate volume for virtual fulls called vf_drive##):
>>
>>      # ls -l drive02/wiki*
>>      -rw-r-----  1 bacula  bacula  5368447065 Feb 17 01:11
>>      drive02/client-aFileVol3000
>>      -rw-r-----  1 bacula  bacula  5368447173 Feb 17 01:19
>>      drive02/client-aFileVol3004
>>      -rw-r-----  1 bacula  bacula  4866003704 Feb 21 01:00
>>      drive02/client-aFileVol3006
>>      -rw-r-----  1 bacula  bacula  2008089223 Feb 26 01:00
>>      drive02/client-aFileVol3134
>>      -rw-r-----  1 bacula  bacula   136824584 Mar  3 01:00
>>      drive02/client-aFileVol3295
>>      -rw-r-----  1 bacula  bacula   236109713 Mar 10 01:00
>>      drive02/client-aFileVol4272
>>      -rw-r-----  1 bacula  bacula   119846566 Mar 15 01:30
>>      drive02/client-aFileVol7651
>>
>>      # ls -l VirtualFull/vf_drive02/wiki*
>>      -rw-r-----  1 bacula  bacula  5368447097 Mar  6 09:07
>>      VirtualFull/vf_drive02/client-aVFVol5720
>>      -rw-r-----  1 bacula  bacula  5368447166 Mar  6 09:08
>>      VirtualFull/vf_drive02/client-aVFVol5722
>>      -rw-r-----  1 bacula  bacula  5368447187 Mar  6 09:10
>>      VirtualFull/vf_drive02/client-aVFVol5725
>>      -rw-r-----  1 bacula  bacula   286220965 Mar  6 09:10
>>      VirtualFull/vf_drive02/client-aVFVol5728
>>
>> So, is there a way for a virtual full to read from itself to generate a
>> new virtual full, without having to move or clone volumes around?
>>
>> Or, is there any advice in general where I can achieve a similar goal?
>> Where I can avoid doing a monthly full, but rely on incrementals and
>> virtual fulls to create new virtual fulls?
>>
>> Thanks!
>>
>> Mike C
> Hi Mike.
>
> You seem to have the same file-pool defined for both jobs. No?
>
> --
> Silver
>
I'll re-verify this, but the VF would never initially run if I specified 
the -vf pool in the Virtual Full job, let me re-test this though and get 
back to you.

Mike C

------------------------------------------------------------------------------
Colocation vs. Managed Hosting
A question and answer guide to determining the best fit
for your organization - today and in the future.
http://p.sf.net/sfu/internap-sfd2d
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users

<Prev in Thread] Current Thread [Next in Thread>