Depends on the amount of data, but I've had great luck with AWS S3fs and glacier for dealing with backups. Cron runs a daily backup snapshot and copies it to s3fs s3 expires the data after 4 months and the space magically re-appears on the disk. Meanwhile it's safely stored in the cloud and glacier is there to help ensure it can eventually be retrieved even if the whole s3 fabric goes down.
Once a month I spin up a dev box and attempt a restore then switch that box to live and move the former live box back into test. Seems to be working so far! On Fri, Mar 21, 2014 at 11:49 AM, Joshua Marsh <[email protected]>wrote: > For a while, we solved this issue by putting our backups on the same > partition as our database. We would rsync, but not delete. Our customers > were great at notifying us when the disk was full. > > > On Fri, Mar 21, 2014 at 10:46 AM, Ryan Simpkins <[email protected] > >wrote: > > > Don't let this be you.... > > > > Then: Backup server became full, backup script silently stopped > > functioning... > > four months later I finally find out when it is time to do a restore. > > > > Now: New fancy backup partition created, updated alerting to notify if it > > happens again, send e-mail to local IT group reminding them to CHECK YOUR > > BACKUPS! > > > > Annual warning complete. > > > > You may now commence the public lashings and/or flame-wars over which > > backup > > solution is superior.... > > > > -Ryan > > > > /* > > PLUG: http://plug.org, #utah on irc.freenode.net > > Unsubscribe: http://plug.org/mailman/options/plug > > Don't fear the penguin. > > */ > > > > /* > PLUG: http://plug.org, #utah on irc.freenode.net > Unsubscribe: http://plug.org/mailman/options/plug > Don't fear the penguin. > */ > /* PLUG: http://plug.org, #utah on irc.freenode.net Unsubscribe: http://plug.org/mailman/options/plug Don't fear the penguin. */
