I have been using OpenMediaVault for quite some time with a single 4TB drive. I had a media directory and torrents directory and I managed all my media using various *arrs. Then I used jdupes to hardlink files in torrents and media to save space and also be able to seed the torrents.
Now recently I bought multiple drives and built a TrueNAS system, I created media and torrent dataset, copied all my data from the 4TB drive to new TrueNAS system. Then I created a Jail to install jdupes in, tried running jdupes torrents and media and realized that linking between datasets is not possible as they are treated as separate storage devices. I found this from here BTW, Reddit - Dive into anything
in the Reddit post, couple of people are talking about fragmentation… I’m not sure what that is or what would be the best solution to avoid that… or what to do at all?
So should I move the data from torrents dataset to media dataset to be able to hardlink as both file/directory need to be in the same dataset?
I also found this, Workload Tuning — OpenZFS documentation
It mentions that you should create a dataset for torrent, and configure the torrent client to move the completed torrents to media dataset… how would this work for seeding?
How should I handle this? I want to space save while also be able to seed the torrents.