On Thursday 30 April 2009 18:29:20 Kevin Keane wrote: > Silver Salonen wrote: > > Hello. > > > > As keeping multiple full backups of the whole data is a very expensive task, I > > think it's wise to minimize the size of full backups. The simple idea for > > doing that is to separate files/folders into "active" and "inactive" ones. > > Active files/folders would then get backed up into multiple full volumes, but > > inactive files would be held in one copy only, eg. by doing only incremental > > backups. If the amounts are smth like 1TB for active files and 2TB for > > inactive files, the save would be noticeable (2TB times the number of full > > backups) :) > > > > The question about Bacula is that is there any way for achieving this without > > scripting filesets? Currently I've done it with simple find-script, but lately > > I tried it on a server having about 1TB of data - when I let the script to > > exclude every old file from the fileset, the incremental job would last about > > 16 hours, even though the backup itself was only 1GB. I guess it's because the > > enormous amount of old files. > > > Is it the script itself that runs that long, or does bacula take that > long to process the output from the script? > > If it is the script itself: maybe you can cache the results of the > script in some form? Have the script generate a text file, and the next > time it runs, read it back in. Or run the script as a cron job > independent of bacula store the output in a file, and then use that. > > If it is bacula itself that takes that long, then you need to find a > better way to exclude these files. > > Is there a way you can move the inactive files into a different > directory tree? If not: think about creating a completely separate > directory for current files. Create links (hard or soft - hard links are > probably better here) from that directory to the original current files. > Exclude the full original data directory from the backup, and back up > only this "shadow".
It was Bacula that just kept thinking about smth for 10 hours and then finally finished the job. The script itself ran within minutes. But I thought of the other-way approach when writing the original e-mail - what if I only include new files, not exclude old files? And when I ran a job with such a fileset, it took only minutes. I'll test this approach and see whether everything is correct etc. -- Silver ------------------------------------------------------------------------------ Register Now & Save for Velocity, the Web Performance & Operations Conference from O'Reilly Media. Velocity features a full day of expert-led, hands-on workshops and two days of sessions from industry leaders in dedicated Performance & Operations tracks. Use code vel09scf and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
