I’m setting up kopia to backup a friend’s Windows PC to my personal ubuntu server via sftp. As his internet upload speed is only 1 Mbps, at his place I copied his 1.2 TB of data on a USB drive and now that I’m back home I’m about to generate a first snapshot from this USB drive to a new repo on my server (via my LAN), then I’ll setup kopia on his PC to open the existing repo on my server via sftp through his ADSL connection, so that the next snapshot will only be incremental.
I’m wondering if for each new incremental snapshot, kopia will [A] scan all of my friend’s 1.2 TB of files and then incrementally sftp the changes (like rsync does), or if [B] there is some sort of cache that allows kopia not to rescan all folders. I assume it will be [A] as kopia is probably not intercepting the system files I/O to figure what has changed since last snapshot…
I’m afraid that for 1.2 TB it will be very long and that it may slow his PC during the scan, so I may split the snapshots into 3 parts of 400 GB each and do one of the 3 snapshots each day.
Any experience on how long it takes to scan 1.2 TB of data before starting the upload of the changes?
I’m also hesitating between creating a batch launched from Windows’ Task Scheduler vs. using kopia-ui (if I understood correctly, if kopia-ui is launched at startup, it takes care of generating the snapshot as per the policy set?). Any recommendation?