Hi folks.
I hope you can help with the best path for recovery from a failed primary backup location.
Here’s my setup:
Primary backup location is an SSD in my server. Backups consist of various folder locations on my ZFS volumes to this SSD.
Total backup size is about 1.7-2TB.
In addition to this location, there are two other storage locations - copy compatible with that primary location. One is local on an external spinning hard drive and the other is BackBlaze B2.
Backups to the primary run hourly. Copies to the two “remotes” run daily. Checks of all three locations run weekly.
Prune jobs run daily keeping all hourly for the first day, daily backups for a week. Weekly backups for a month and monthly backups for a year.
This has been running successfully since 2020 and has saved my with successful restores from local primary and secondary locations on multiple occasions.
Yesterday, my primary backup location SSD appears to have not survived a reboot of my server. It may be recoverable, but I have not tried anything regarding that yet.
The drive appears briefly and responds to SMART inquiries for a minute or two after being powered down.
I’m not in imminent danger of data loss because the actual source drives are good (Mirrored spinning disks) the most recent check of the copies of the backups both were clean (5 days ago) and the most recent copy of the primary to the secondary copies succeeded (yesterday morning).
But, since my primary backup location is down, I am not currently getting new backups. We don’t generate a ton of new data on our home server, but I definitely want to address this ASAP.
So my question is what is my best approach?
The way I see it I can either:
a) rclone or rsync the local USB external copy of the storage to a new internal drive to replace the failed one and then adjust my storage and backup and copy definitions to use that new copy as the primary location
or
b) change the primary backup location to be the local external USB drive and then use it for the source of the copy to B2 and when I replace the failed primary copy back to that location as well. Once this process is done I can then repoint the primary backup location to the new copy on the internal drive and repoint the copies back to that as source and the external and BackBlaze as destinations.
So,
will either of these approaches work? If so is there any material difference between them?
I guess my main questions are
Are copy compatible storages able to be either the source of destiation for copies from other storages that are compatible with them?
When the primary compatible storage has failed, can I create another compatible storage location from one of “its” compatible locations.
Am I missing anything?
Thanks in advance for any help.
Ted