feature request: check lockfile on remote systems

[ian-kelling]

I can imagine timing issues that would make this not work perfectly, but it would still be useful.

[digint]

The reason why I did not implement remote locking yet is that I don't think it is really needed:

If you are fetching from remote, the snapshots are either already created, or, if not already created, they will be transferred in the next call. Now there is only "valid-ish" reason you would want that: you might want to assure that all snapshots are transferred from the current btrbk run on target. But then again, with a lock, you would just be waiting for the target to finish the job instead of just taking what is already available. If you care about availability, just increase the interval you fetch from the remote (in the worst case btrbk would just do nothing).

Same applies for pushing to a remote host.