qBittorrent official forums

I come here to vent and ask your opinion about something weird that's been happening with my qBittorrent. Let me preface saying that at first I thought it was a plain instance of metadata not downloading. But after scouring the bug trackers and forums I've realized nothing seems to describe what I was seeing. And then it all went to hell.

I was running v4.4.2 of qBittorrent, run as a service through NSSM. Almost all the torrents I got through magnet links. Then the files are moved to another hard drive and served to a local network via a different service that picks the completed torrent's files. After some headaches this started running fine until about two months ago when I noticed a huge chunk of torrents stuck retrieving metadata.

The funny things is, the content of said torrents was on the server. Somehow, the files downloaded correctly(?), even though qB claims to not know what the contents of the torrents were. I didn't notice earlier because the files were being moved correctly afterwards. I don't know what was going on.

I have only a handful of sources I download from, and I've noticed that online magnet to .torrent converters have issues resolving the metadata of some of the torrents as well. Nevertheless, given enough time, some new magnet links added to qB manage to download the files while not downloading the metadata. Others, I don't know why, don't download anything despite having hundreds of seeds.

I suspected there's some DHT fuckery going on, and even asked (with no response) the admins of the sources I download from. It could also be an error with qB local database(?). I didn't know what to do, it's just seemed completely random whether a magnet link will download correctly or not and I could never know what is the status of anything. As a result, I couldn't seed these faulty torrents either.

So, reading all the problems with libtorrent 2 and v.4.4.x I decided to downgrade to v4.3.9. Stopped all torrents, uninstalled, installed the downgraded version. Checked that everything was working properly, so it seemed. All the files that were downloading continued where they were, and everything else was as it were. I decided to let it run for a while to see if this version could correctly retrieve the missing metadata and sort the faulty torrents.

I went 2 hours to do something else in the house and when I come back, over 40 torrents had their data wiped out. Nearly 250GB of data that was already downloaded and allocated correctly completely gone. Somehow qB thought the already completed downloads, weren't and so decided to erase everything in the hard drive.

I hate that I'm stuck with qB. I don't want to use torrents, but my internet connection is so slow that this is the only reliable way of fetching large data sets reliably. I hate that qB is the best FOSS torrent software and still is so obscure, quirky, finicky and unpredictable. I don't know what this software is doing half the time. And somehow all the other alternatives are somehow worse.

I'm just so tired of so many problems. I wish I knew netcode programming to fork it and contribute but I don't have the time or the knowledge. I'm just worn out that every quarter some new unintended behavior pops up.

I'm practically back at square one with almost all lost.

Sorry about that, but if you want to play on the safe side you can archive your torrents. Just be aware that the info-hash is really all you need about a torrent, the magnet link is just built from it and the torrent file can (if sources exist) be downloaded from the DHT.
A simple solution to that is use webtorrent-cli which you can find on github and it's an npm package (npm install webtorrent-cli -g)

webtorrent downloadmeta <torrent-ids...> Download metadata of torrent

The argument can be just an info-hash, and it will save a torrent file. You can use the webtorrent-cli to also download torrents. Depending on your automation scenarios, there are many options.

Hope you managed to download your data back