droolio

If qBittorrent/qb-nox is bound to your VPN interface, then 1) your VPN needs to support port forwarding, and 2) forwarding a port on your router is pointless and unnecessary. Your only way around it is to switch VPN or don't use VPN and then port forward.

Actually, ufw has its own separate issue you may need to deal with. (Or bind ports to localhost/127.0.0.1 as others have stated.)

The next best alternative would be BiglyBT's Swarm Merging feature (which works similarly, and amazingly well on v1 torrents considering it only stores a precise file size instead of a hash in Vuze/Bigly's own DHT). I've been able to 'complete' numerous separate torrents where availability was <1.

BiglyBT already supports v2 but dunno if Swarm Merging works with such torrents yet.

Thank you for posting this, hadn't heard of it before.

Yes, I also work in IT.

The paid GUI version is extremely cautious on the auto-updates (it's basically a wrapper for the CLI) - perhaps a bit too cautious. The free CLI version is also very cautious about making sure your backup storage doesn't break.

For example, they recently added zstd encryption, yet existing storages stay on lz4 unless you force it - and even then, the two compression methods can exist in the same backup destination. It's extremely robust in that regard (to the point that if you started forcing zstd compression, or created a new zstd backup destination, you can use the newest CLI to copy data to the older lz4 method and revert - just as an example). And of course you can compile it yourself years from now.

The licence is pretty clear - the CLI version is entirely free for personal use (commercial use requires a licence, and the GUI is optional). If you don't like the licence, that's fine, but it's hardly 'disingenuous' when it is free for personal use, and has been for many years.

Wouldn't you be on CGNAT though? How are they blocking it - at the DNS level? Have you tried a CNAME record that points your own domain to the actual duckdns domain? Just curious how/why they might be doing this.

If they're supposed to be binary-identical data (same file checksums), you can use BiglyBTs Swarm Merging feature - without manually copying (which isn't as reliable due to the start/end of the files not bordering on the chunk boundaries.

If they've been modified in any way though, this won't work. However, you might be able to use its Swarm Discovery to find other torrents with the same data and complete with Swarm Merging.

It's no different than any other client, and looks like the torrent is healthy so should just work.

As with all BT clients, you get better connectivity by forwarding the designated port. Does it say NAT OK at the bottom of the window?

Yea qBittorrent, chosen as was meant to be lightweight so runs in a docker container on a Raspberry Pi 4 (but often eats all 8GB available RAM - could be just a problem with the docker imagine tho.) Haven't used Transmission so can't compare.

Compared to the memory footprint of web browsers, a Java BT app is pretty tame on modern computers these days. Nor do you have to faff around with installing JRE manually. It just works.

Resource usage is pretty good tho and it can handle hundreds of torrents with ease.