Thank you! I've done some more digging with this, and it seems that doing a backup to disk is saner for my scenario. For now, I've setup an Always Incremental backup with the Tape Drive as the long-term storage, and will look into the VirtualFull solution (which seems to require a different job per actual backup job).
Michael On Tue, Feb 5, 2019 at 10:21 AM Douglas K. Rand <[email protected]> wrote: > On 2/4/19 8:46 PM, 'Michael Stum' via bareos-users wrote: > > I have implemented a backup that does Full, Differential, and Incremental > > backups to tape, and all that is working. > > > > However, I would like to always keep a Full Backup of each client around > on > > disk. And not just the last Full backup, but a VirtualFull that gets > > updated with every Differential and Incremental backup as well. > > > > I looked at copy jobs, but if I understand this correctly, I can't copy > an > > Incremental to a VirtualFull backup? I know I could backup each client > > twice a day, once to tape with Full/Incr/Diff and once to disk with > > VirtualFull, but that seems like a waste. > > > > One option that I can see is to forgo the idea of VirtualFull and have > Copy > > jobs that copy each Full/Diff/Incr backup and set retention policies to > be > > short enough to only keep the last Full + any Incr/Diff backups since > then > > around. > > > > Is that the proper way, or is there some mechanism? (Basically, trying to > > do hot and cold backups) > > I'm not sure that there is a proper way to do this. Just a way that works > for you. > > What we do is kind of the opposite of what you are asking. :) Our normal > backups are stored on disk, and then we do a virtual full backup to tape. > We > used to do a full/differential/incremental backup scheme, but have since > switched to always-incremental. But I don't think that matters for this. > > For us the on-disk backups are what we use all the time, but the tapes > (stored > off-site) are for the Oh Shit! moments. > > I don't see why you couldn't reverse that, and build on-disk virtual full > backups from the tapes. But I'd think that having to read all of the > backups > (most recent full, most recent differential, and all incrementals since > the > last differential) would be expensive from tape. > > Here is an outline of our offsite virtual full job defs: > > job defs { > name = "offsite" > type = backup > level = virtual full > schedule = "offsite" > # Gotta turn "virtual full" jobs into Archive jobs so we don't > # try to use the backup for the next "virtual full" job. > run script { > console = "update jobid=%i jobtype=A" > runs when = After > runs on client = No > } > file set = "standard" > accurate = yes > messages = "standard" > write bootstrap = "/var/db/bareos/bootstrap/%n.%i.bsr" > pool = "full" > full backup pool = "full" > differential backup pool = "differential" > incremental backup pool = "incremental" > spool attributes = yes > spool data = yes > next pool = "offsite" > } > > You won't want the "spool data = yes" part if you are building the virtual > full backup on disk. > > -- > You received this message because you are subscribed to a topic in the > Google Groups "bareos-users" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/bareos-users/1m4-kbveSXI/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > [email protected]. > To post to this group, send email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "bareos-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. For more options, visit https://groups.google.com/d/optout.
