Saving bandwidth for my desktopp

My main desktop (linux) is on wifi. That’s a pretty precious commodity around here. I plan to back that up to a small linux server running a kopia server with a local repository. That server has a wired connection to my router.

My question is, when I run sync-to on my desktop to sync that repository to b2, will that use much wifi? Or will the traffic be from the server to b2 (as I hope)?

If you running dedicated, separated kopia server and backing up to it your desktop then why you using sync-to on desktop instead on running sync-to from kopia server?

If you use sync-to from your desktop, the data will run through your desktop and therefore your WiFi. If you use sync-to from your server, the data will run through your server and therefore not your WiFi.

Well… that’s a tough nut to crack, I suppose. Backing up to a local repo saves you from a lot of small traffic, when kopia checks for manifests and such. The following sync-to will be more of a streamed connection where kopia will check either source and target repo, compute the fragments which needs uploading and do it.

I could see how this is a little bit less involved for the Wifi connection. You will especially save time, when running the snapshot against the local repo. But, if it will be actually saving a lot of bandwidth, I don’t know.

Because I’m backing up my desktop, so that’s where the primary cron jobs for backup will run. It’s the only place that it’s easy to know “I just finished a backup, so sync to b2”

I suppose if I don’t care about timing to b2, I could just run the sync-to on the server, and some snapshots might take longer to get sync’ed to the cloud. I’m assuming if a sync-to is run on the server while a backup is running from desktop to server, that won’t cause a problem, other than maybe that backup not getting sync’ed until next time.

Did you check session list command ?
Also, kopia has actions, that you can use as signaling mechanism by writing some script and notify server - “backup has been done”, so on the server side you can run then server pause (just to be sure) and fire up sync-to as a client

I’m still just doing test backups to my local drive, and I’m still in the planning phase. But my external drive for the server arrived today, so I’ll start setting up the real process tomorrow.

Thanks for the tips.