Starting my journey into the "Arrs" and getting comfortable with their various workflows; KRONOS put on a schedule to preserve my hearing
Finally settling on a file-folder organizational convention for some automations. More testing needed though!
SABnzbd and Sonarr/Radarr/Lidarr Integration
SABnzbd is my Usenet downloader of choice. I spun up a Debian 12 LXC (privileged this time so I could mount NFS shares more easily) and installed SABnzbd from backports. Then I used a simple systemd file that runs SABnzbd as a non-root user, which I find cleaner than the older init scripts. Once I confirmed it was working on port 8080, I integrated it with Sonarr:
- Sonarr (TV): Host is
sabnzbd.internal
, port8080
, plus the SABnzbd API key. - Category: Named it "tv," so any TV episodes are placed in
/mnt/ISOs/servarr/usenet/tv
by default.
Sonarr automatically scans that folder and moves episodes to the final "TV Shows" library. I tested by importing older shows and seeing if they'd get recognized. So far, so good. Manual import also worked well: Sonarr re-labeled episodes and slotted them into the right show/season subfolder.
Radarr is basically the movie-oriented twin of Sonarr, so the setup was
nearly identical. I gave it a separate folder:
/mnt/ISOs/servarr/usenet/film
. One note is that Radarr can be more
finicky about movie "quality profiles," e.g., distinguishing between
"Remux," "WebDL," and "Full Bluray." I prefer limiting my maximum file
size so I don't accidentally grab 50GB ISO images. Once that was set,
Radarr integrated seamlessly with SABnzbd as well.
I'll write a more specific post once I have used it long enough to form an opinion.
Music management is trickier---metadata sources for albums and tracks
can be inconsistent. Nonetheless, I installed Lidarr the same way,
hooking into the same SABnzbd instance and gave it its own subfolder:
/mnt/ISOs/servarr/usenet/music
. I'm less confident in the availability
of older or obscure music and the metdata quality on Usenet or private
trackers is not good but it's a start.
Jellyseer: A Front-End for Requests
As a final flourish, I installed Jellyseer in a Docker container to let family or friends request content without needing to contact me directly. Jellyseer is basically a user-friendly interface: you see a show or movie you want, click "request," and the system lines it up in the background. It also tells you what is already requested, what is on the server, and offers suggestions about what to watch next.
I just mapped the container's config directory to
/srv/jellyseer/config
. Once I set up my subdomain in HAProxy and added
a Jellyfin API key to allow people to use their existing JF credentials
it was good to go. Despite having full access to everyone's viewing
history I make a point of never mentioning what I saw them watch, or
when they accessed the server. I check the stats once in a while to make
sure that if they are watching a season of a show that isn't fully
downloaded that I add the rest of it quietly behind the scenes. To
complete the illusion of privacy I can now let Jellyseer be a neutral
place for new requests.
Housekeeping
Syncoid: The mass-downloading was causing KRONOS to run for many hours at a time. I had to throttle the downloads and switch KRONOS to wake up right as I left for work so that I wouldn't have to listen to it howl all day.