From: Shawn P. <red...@co...> - 2010-01-07 00:03:16
|
I run BackupPC in a OpenVZ VM so I can back up the whole machine easily, it makes mirroring it very easy. On Wed, Jan 6, 2010 at 4:36 PM, Adam Goryachev < mai...@we...> wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > dstahl wrote: > > My apologies if this has been answered before. I tried searching > > around the forums and could not find a straightforward answer to this > > question. Our set up is one backuppc server on the lan backing up > > multiple servers, we are then off-siting all the data to another > > server via rsync. This is working fine. However when I got to the > > point of testing our restores from the off-site location, that is > > where things get a bit confusing. > > Only because you haven't identified your requirements precisely > enough... perhaps? Let's see... > > > First I tried installing backuppc > > on the offsite server and doing a restore from the web-gui. big > > mistake. not only did it erase some of the full and incremental > > backups because I forgot to set the retention periods as the same on > > both servers, > > You mean your backup server wasn't backing up your config from the live > server? This is something to learn from, ie, this is WHY we TEST things. > So we can find out we forgot to sync the config before we really need to > do a remote restore... > > > but also entirely screws up rsync script, since the > > offsite server's backuppc install modified files in the data > > directories, > > I would think the *install* of backuppc doesn't modify any content of > the data directories. However, running the newly installed backuppc > would eventually run the nightly script, which would expire (delete) a > heap of old backups (because your config was not what you wanted). Once > those backups are expired, the nightly script would then prune the > pool/cpool and remove all the files that are no longer needed by your > backups, in the process renaming some pool files where files with > colisions have been removed etc... > > > therefore causing almost a full rsync each time instead > > of just new and modified rsyncs. > > So, if you properly configured your backup backuppc, then this issue > shouldn't occur. > > The second problem with what you are testing, is you forgot to explain > under what scenario you would need to restore files from your remote > backuppc server, and yet still have all your data stored on your local > backuppc server to rsync nightly? ie, if your local backuppc server is > OK, why would you need/want to restore from your remote backuppc? > > > Next I tried the command-line > > restore. putting this all in a tar seems crazy since it is almost 3TB > > of data. That would mean I would need 6TB of space, which I don't > > have. > > Why do you need 6TB of space? Do you mean you need 3TB to store your > backups, and another 3TB for a single copy of all your data (the restore > tar)? Or is there a temp file created somewhere that needs 3TB? > > > I tried piping the command like so: BackupPC_tarCreate -h > > myserver -n -1 -s '/server/shares' / > pm.tar However it gives me out > > of memory errors. (this was done on a lab server. not sure if real > > server with more memory will give same error) > > > > So here is my condensed question. Given my scenario, Backuppc to > > offsite server via rsync. What is the best way to do a full (and > > partial) restore from the offsite server? > > Better to define your requirements first... > I would suggest a number of options: > 1) Attach an external (esata) drive to your remote backuppc server, do a > restore (from the gui, or cmdline) with the destination set to the > external drive, then take the drive to your LAN (or recovery location) > and use as needed. > > 2) Use rsync in reverse to copy the remote backuppc pool back to your > primary backuppc server (assuming there was some minor issue with it as > opposed to total data loss). > > 3) Use dd to copy your remote backuppc to your primary backuppc either > with a esata drive or over the net (with nc or ssh etc) > > These are a couple of options, but in my limited experience, you would > only be looking to your remote backuppc system: > 1) Your local backuppc system has died (stolen, RAID issue, multiple HDD > failure, etc) > 2) Your local site has been destroyed (fire, theft, natural disaster, etc) > > There is perhaps one scenario you are not protecting against, which is > sabotage/hacking/etc. ie, an Internal or external agent manages to rm > - -rf your local backup machine and/or one or more servers on the LAN. > Will your remote rsync replicate this data destruction. leaving you with > no backups? ie, the malicious person does the deed at 2am, and your > rsync runs at 3am... will someone notice in time to stop the rsync and > save your final copy of the data? > > I'm not trying to be negative, or suggest that you haven't thought about > any of the above, you haven't provided enough information to know that. > However, hopefully, it will provide some information that you or someone > else has not thought about/considered... > > Regards, > Adam > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.9 (GNU/Linux) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org > > iEYEARECAAYFAktFHmMACgkQGyoxogrTyiVleACeMVxvOXYbUI0Kl7eXv7SQPRn6 > 5L0AoMwfxIq2qo6fQ+E2lPFgjLEfB8FN > =lWHA > -----END PGP SIGNATURE----- > > > ------------------------------------------------------------------------------ > This SF.Net email is sponsored by the Verizon Developer Community > Take advantage of Verizon's best-in-class app development support > A streamlined, 14 day to market process makes app distribution fast and > easy > Join now and get one step closer to millions of Verizon customers > http://p.sf.net/sfu/verizon-dev2dev > _______________________________________________ > BackupPC-users mailing list > Bac...@li... > List: https://lists.sourceforge.net/lists/listinfo/backuppc-users > Wiki: http://backuppc.wiki.sourceforge.net > Project: http://backuppc.sourceforge.net/ > |