I have been using rclone mount and rclone docker for some time and I have some queries about the VFS cache and was hoping they could be clarified.
I mainly use rclone → crypt → sshfs and know that sshfs has a number of error codes that can be returned (file path too long for instance is a permanent error). Rclone takes the file operation (before validating if the destination filesystem would accept it, admittedly, sometimes it won’t know), commits it to VFS cache and then tried to get the file to SSHFS. If that fails for some reason, the file appears to be stuck in the vfs cache forever until it is manually resolved, or the cache is cleared and the data lost.
What happens if I disable the VFS cache completely, does that mean that the error code will be returned, and the operation will not complete, or will the file be taken, and deleted ?
I would guess this could occur in any situation where the parameters of the destination system are not completely compatible with the local file system.
Is there a configuration setting where after 5 attempts the file would be returned to where it was, although I would guess that would be just as difficult and prone to error if another file with the same name was present and modified.
Sorry if this is a daft question, I can’t get it straight in my head how this can work safely.