From: Naufal Z. <nau...@gm...> - 2007-02-20 19:16:59
|
Martin, thanks for your reply. I am still not sure about the exact problem. Please correect me if I am wrong on following: As the example states, it looks like that there has to be one schedule ( as compared to 3 seperate I have) which will spepcify the level of the backup i.e. full, differential and incremental and the timings. So I should have one schedule which will do all the scheduling rather than having three. For the job definitions, from your comments and the example my understanding is that I will have only three jobs for my three systems instead of having seperate full, differential and incremental jobs with different schedule for each system. Each job resource will have the backup pools defined with that single schedule directive name I had defined in schedule section. As a result the system will take a full backup on the full backup pool, and the differential backup will reference the same full backup for the differntial backup instead of reverting to its own full backup and similalry the daily backup will reference the same full backup instead of reverting to its own full backup. thank you so much for all your help. Regards >From your comments my understanding is that On 2/20/07, Martin Simmons <ma...@li...> wrote: > > >>>>> On Mon, 19 Feb 2007 14:46:07 -0500, Naufal Zamir said: > > > > Hi Martin, > > > > thanks for your reply. i really appreciate it. I don;t think the job > > retentions should be a problem, however I did not get the level options. > I > > tried using the same name for the jobs, but check conf gave me an error > > saying that no two jobs can have the same name, though the backup level > was > > different.I am however attaching the conf file for your review. I'll be > > gratefull if you can pin point the mistake, cos this seems to be the > last > > hurdle in my implementation! > > Please always CC the list in replies. > > Sorry, I wasn't clear about the job definitions. You need 1 job > definition > for each set of files and client that you want to back up, e.g. named > something like "Tracfire", "Tracweb" and "Tracsun". The job definition > can > use the Full Backup Pool, Incremental Backup Pool and Differential Backup > Pool > options to specify the pools for each level. > > See http://www.bacula.org/rel-manual/Automated_Disk_Backup.html for an > example. > > __Martin > > > > > > > > On 2/19/07, Martin Simmons <ma...@li...> wrote: > > > > > > >>>>> On Mon, 19 Feb 2007 11:02:35 -0500, Naufal Zamir said: > > > > > > > > I am running bacula 1.38.5. > > > > > > > > I want to know if the incremental and differential backups can tale > a > > > fuill > > > > backup from a different pool as their reference point. > > > > > > Yes. > > > > > > > > > > Currently, I am using a full backup pool for the full backups, and > since > > > it > > > > takes pretty long for all the system's full backup I do it on the > > > weekend. > > > > After the weekend when bacula is supposed to do an incremental or a > > > > differential, it does not find any backup for that particular job > and > > > > reverts to full backup. what I want to do is that differential and > > > > incremental should not revert to full backup and take the full > backup > > > from > > > > other pool as their reference. Otherwise since it take 3 days for > one > > > full > > > > backup of all three systems, whole cycle disrupts. > > > > > > There are two common mistakes: > > > > > > 1) How did you define your job? To make differential/incremental > work, > > > you > > > must use the level option when running the job with the same name > as > > > the > > > full job. I.e. one job with a schedule that contains more than one > run > > > option. > > > > > > 2) What are the retention times for the job? If they are too short, > then > > > Bacula might prune the job records for the full run before it runs > the > > > differential/incremental. > > > > > > __Martin > > > > |