Very timely from my perspective. My backup strategy was (MacOS):
- primary catalog on internal HD, backed up by LR to Dropbox and Time Machine to a pair of external drives
- primary image storage on external HD, local backup via Time Machine as above, offsite backup to CrashPlan.
- Additional copy of imported files on alternate external HD, backed up to Time Machine and CrashPlan
- Additional copy of Selects Raws to Dropbox, and JPEGs to Flickr
Last Tuesday morning the external drive housing the primary image storage failed - refused to power up at all. I immediately ordered a new hard drive, and went to Time Machine to retrieve the local copy. For reasons I still don't understand, that drive was missing from Time Machine (probably user error). However, CrashPlan appeared to have all 1.3TB of data, so I began restoring to my internal HD.
By early afternoon, the new external drive arrived and was installed, and I redirected the restores to that location.
By Thursday evening, all 1.3TB were restored from CrashPlan and LR was repointed at their new location.
By Saturday, CrashPlan finished the backup of the new HD, probably by noting that the new files being backed up were already at their site. Time Machine now appears to have all of the image files as well.
So ...
- multiple layers of backup, using multiple technologies. Technologies fail, users make errors, etc.
- onsite and offsite - your onsite copies could burn, you offsite provider could fail
- check that the backups exist and you can actually restore by changing or deleting a primary file and recovering it from each backup