I use dar[0] to create incremental, compressed, encrypted backup archives, and then upload those to a GCS bucket. GCS is dirt cheap even for terabytes of data, so as added bonus this gives me a
history of all my files (eg, i still have backup archives from 2013). Both calls wrapped into a shell script that is regularly scheduled via a cronjob on my local server. Within my network i sync between laptop, pc and server via syncthing. Works well for me, even though it doesn't handle file renames gracefully.