What is the best way to backup locally as well as to a cloud provider with minimum overhead.
There are different types of overhead you could refer to so it’s hard to answer, but my general advice is to simply set up a second backup job to your second repository. It doesn’t have to be more complicated than that, and has its pros.
By overhead, I mean running two concurrent backups. I know ultimately that’s what has to happen, but having them no collide and have issues or happen at the same time so performance suffers. Having an offsite and having a local copy is critical.
You can back up the same files from the same source at the same time with restic, no problem, I’ve done so myself for years. Is there an actual problem you would have with running two backup jobs on the client (at the same time or not)?
Copy local backup with rclone
--immutable Do not modify files, fail if existing files have been modified
to remote location.
It’s basically what I did with Windows, using two backup solutions. I see backing up one, then syncing off site mentioned a lot but that isn’t a viable option as that has more risk.
I backup to a local disk, then I mirror the local backup to AWS S3 using the AWS CLI v2 command “aws s3 sync”.
I’d like to avoid that because any problem with your local backup is now mirroed to your offsite one defeating a lot of the purpose of it. I ended up just writing two scripts and running two backups, one local, then one offsite. I also have things I don’t want sent offsite, for example my Steam game library, I like to have this local so I can restore them easily, but I don’t want to send hundreds of gb offsite for someting I can easily reinstall.
That sounds like a good approach. I have half a dozen different restic repositories. One is deliberately mirrored to S3, a few stay local, a few are on external disks which usually live offsite.
I’ve been using Restic for a few years and haven’t had any problems yet. I do restore tests every six months to check it’s all working ok.
That option seems the best to me, a backup for each repository. They are independent of each other and if one fails the other can save you.
Any tips or tricks on how to name the rclone repositories?
Backup, Backup2, does not seem like a good strategy because then if you want to know where each one points, editing the repository only shows you the credentials.
Maybe something like
Or is there a way to add a description to a repository?
Like in programming, using long descriptive names is better than short cryptic ones.
I use folder names like “Photos” , “Daily Backup”, Offsite Backup Work", “Long Term Archive”. Batch files / scripts are similarly named.
Thanks, I have finally opted for compound names: computer, backup type and repository location