I have a lot of jobs running, and if a source machine is offline, they generate errors in the mount script (this is by design). Then in the morning when I check Duplicati, I have as many as ten “Error” messages in the GUI that I have to manually dismiss. The relatively slow GUI makes this somewhat annoying because the “Dismiss” buttons are a little slow to register (maybe AJAX?):
My use case is a bit unique. I have Duplicati running on a central server, backing up machines from my LAN to various onsite and cloud destinations. I want to manage the backup from one place. For laptops that are on the LAN, I have a script that mounts the user directory to a local mount point, and if it fails, this means the machine is offline/unavailable/sleeping. My pre-backup script will give an exit code in that case (I chose 3 so I know right away that it was simply a mount failure).
I suppose if there were a parameter that said “gracefully-exit-from-pre-backup-required-script-exit-code” and had a field where I could set the exit code of my choosing, such that if the exit code matched, the backup would silently fail with no errors/warnings, that might help, but at the same time I don’t necessarily mind getting a warning when that machine doesn’t back up. If it happens many days in a row then I make a point to ensure it is online from time to time so it does get backed up at least once a week.
I do wish Duplicati had more of a client/server architecture, but I know that isn’t the mission of the developers so I’m trying to fit something into an arrangement that isn’t what it was designed for. The only reason this particular issue was annoying was that I’d get an error for each destination. I know there has been talk of allowing for multiple destinations for one backup job, but for now this will work nicely.
Thanks for the assist with --allow-missing-source. For those that might care how it’s defined on the parameters page:
Use this option to continue even if some source entries are missing.
Default value: “false”
I wonder how defaulting that ON would play out if / when drastic change alerts get implemented.
Thanks for the consideration on that.
I expect you’ve already seen it, but depending on what “server” functionality you’re expecting there are two or three tools others are developing to centralize reporting and possibly even some centralized control tools.