Educational: Force concurrent backup errors/inconsistencies/whatever

my use case is that I have several jobs on the same host backing up the same folder to the same restic repo. I.e., two snapshots can be told apart from time and tags only.

These jobs are meant to be run sequentially, but I cannot guarantee that. I.e., two or more jobs could run in parallel.

Am I in trouble, or will restic handle that situation fine? (I only read that concurrent backup jobs from different hosts are ok, but could not find info on my particular case.)

Thanks for making restic and for any comment clarifying my question.

I’m not aware of any issues, restic should handle this smoothly. But the best thing is to try it out and see what happens :wink:

Are you referring to this post? I think this is caused by stash, not restic.

Thank you for the swift response! I was not referring to that post, however it is a good source of additional info.

May I ask a follow up question. My jobs change contents of the source folder, e.g. by moving files or by changing files. Is it harmful if one job changes source folder contents while the other job is in the middle of a restic backup? I.e., should I establish sequential processing of my jobs?

Big thx again.

restic will throw a warning if the preliminary scan found a file or directory which can not be found during the actual backup task. The backup job does continue in this case but will of course not contain this file/directory. If this is a problem you can add filesystem snapshots to your backup jobs.

Note that you might wind up with some duplicate data in the repository if both restic processes encounter the same new data at the same time. This is actually a fairly common occurrence and isn’t an error, only a temporary inefficiency. A future restic prune operation will remove the duplicates.