My qbittorrent downloads are stuck downloading metadata

Posted by Itskorn

My question: any interface all addresses SparkBox AI replied: Yep, those bindings are correct. DNS resolves, the VPN is up, and qBit isn’t pinned to a dead interface. Check one last qBit setting: Settings → Connection → Proxy Server It should be: text Type: None If a proxy is selected, set it to None, save, then restart the media module: bash sudo sparkbox restart media A stale proxy config can produce exactly this kind of tracker/metadata nonsense even while DNS lookups work. If it’s already None, we’re past the normal config fixes. Click Share to demox below with: - DHT remains 0 - Tracker statuses: Host not found and Operation not permitted - getent ahostsv4 exodus.desync.com resolves successfully - qBit binding is Any interface / All addresses - Proxy is None if confirmed Tom needs the full qBit config and media wiring on this one. What I still need help with: (describe what didn't work or what's still broken)

3 replies

Anthony wrote:

Thanks — "Any interface / All addresses" is one useful result. I need to correct the earlier conclusion: it does not by itself verify DNS or the VPN, so please do not change the proxy or restart anything yet. In qBittorrent, select one affected download and open its Trackers view. Reply with: 1. whether any tracker says Working and shows seeds or peers (you can redact tracker URLs or passkeys); and 2. which platform hosts SparkBox — Docker Desktop, Windows with WSL Docker Engine, or NAS/Linux — plus the installed SparkBox version. That will distinguish tracker reachability from a torrent with no available peers and avoids a speculative repair.

Itskorn wrote:

DHT, PeX, and LSD are the only working trackers. i am using raspberry pi os with Sparkbox version v1.6.670

Chris wrote:

Thanks - that answers it clearly. DHT, PeX, and LSD don't talk to a tracker at all (DHT is trackerless peer discovery, PeX comes from peers you're already connected to, LSD is local-network broadcast), so "only those three are active" means none of your actual trackers are connecting - it's not a shortage of peers. One thing worth flagging honestly: SparkBox officially targets x8664 servers, and Raspberry Pi OS/ARM isn't an officially supported platform, so some of the VPN container's networking behavior hasn't been verified there end-to-end. That may or may not be related to what you're seeing - I don't want to guess either way. To find out whether the VPN container itself is blocking the tracker traffic before it leaves the box: can you try adding/rechecking a torrent, then run sudo docker logs sb-gluetun --tail 100 right after, and paste any lines that mention DROP, REJECT, firewall, or permission? That tells us whether gluetun is blocking the announce, before we touch any settings.