Understood that restic can’t directly push data to the Backblaze backup service and can push directly to B2. However B2 is cost prohibitive for the amount of data I am looking to store (~7TB+ without any versioning).
I am not looking for restic to push data directly to Backblaze backup service though. this is my planned configuration:
Linux Server <---- restic via SSH -----> Windows Server <---- Backblaze client -----> Backblaze backup service
Restic will be used on the Linux server to send backups, manage snapshots etc to the restic repo on my Windows server. From there the Backblaze client can backup all of the restic raw repo files (config/data/index/keys/snapshot folders).
If my Linux server ever needs to be restored, then I can use restic to restore from the repository on the Windows server. If the Windows server dies, I can use restic to redo the backup.
Backblaze will only be used for disaster recovery in the case that both servers are lost at the same time. I know i won’t be able to directly restore the data using restic. I will have to use Backblaze to restore all of the
restic repo files to the Windows server first, and then I will be able to use restic to restore the actual files to the Linux server.
My only concern is how restic stores snapshot information. If a new snapshot has been created on the Windows server, but Backblaze hasn’t finished uploading all of the new repo files I want to make sure that the repo files stored in Backblaze are valid (in particular i’m worried about files in the index folder).