So now that we’ve setup our arr-stack and are able to download stuff either via usenet or torrents we can actually set up something to consume our media on!

I’ve used Plex in the past and while it is a much more polished experience than jellyfin, the decisions made by the plex team are not really confidence inspiring when it comes to privacy. So i’ve switched over to jellyfin and i am more than happy with it!

Jellyseer on the other hand is a much more polished UI for Sonarr and Radarr. It connects to your jellyfin instance to chech what is available and it connects to your sonarr and radarr instance to tell them what you want to download. It’s super simple, has an integrated browser for all shows and movies for you to find and looks really frigging good. This is also amazing if you “offer your services” to friends and family, as this is a super easy to understand website for your other users to “request” shows and movies. You can configure if some users requests should automatically be approved or if you want to manually click yes and no for each request.

So let’s get going!

0. Preamble

All these guides assume you have basic knowledge about docker & docker compose. To quickly summarize the methods here:

You either use a webui to manage your docker-compose like Dockge or you create a folder and a docker-compose.yaml for each service listed here and copy the contents of each of them into that docker-compose.yaml; IMPORTANT: You HAVE to name them docker-compose.yaml exactly like that. Do NOT name them radarr.yaml for example. After you created a docker-compose.yaml and filled it with the contents here, you simply run docker compose up -d

Since neither Jellyfin nor Jellyseer actually download some “problematic files” there is no need to run them through any VPN. Of course it doesn’t hurt to route them through the same VPN you have for sonarr / radarr and the likes.

I will provide two seperate docker-compose.yaml for the Jellyfin installation. One with hardware accelerated transcoding for NVidia cards, and one that has no hw-accel.


1. Folder Structure

This is based entirely on my previous posts on how to set up your arr-stack (usenet or torrents. If you don’t use an arr-stack or already have one up and running with a different folder structure you’re on your own here.

Since we already established this folder structure we can simply add two more folders for jellyfin and jellyseer in the configs tree:

arr-stack (for this example let's say it's located in /mnt/arr-stack)
	|---- media
	    |---- movies
	    |---- shows
	    |---- music
	    |---- etc.
	|---- configs
	    |---- radarr
	    |---- sonarr
	    |---- qbittorrent
	    |---- jellyfin
            |---- jellyseer
	    |---- etc.
	|---- torrents
	    |---- complete

also, if you have a fast SSD in your system with enough free storage you can also create a folder there for your jellyfin cache. Let’s assume it is on /mnt/fastssd/jellyfin/cache This can make a drastic difference in performance and buffering.


2. Docker-Compose Files for each Service

2.1.1 Jellyfin without hardware accelerated transcoding

services:
  jellyfin:
    image: jellyfin/jellyfin
    container_name: jellyfin
    user: 1000:1000
    network_mode: host
    volumes:
      - /mnt/arr-stack:/mnt/arr-stack
      - /mnt/arr-stack/configs/jellyfin:/config
      - /mnt/fastssd/jellyfin/cache:/cache
    restart: unless-stopped

2.1.2 Jellyfin WITH hardware accelerated transcoding (NVidia)

services:
  jellyfin:
    image: jellyfin/jellyfin
    container_name: jellyfin
    user: 1000:1000
    network_mode: host
    volumes:
      - /mnt/arr-stack:/mnt/arr-stack
      - /mnt/arr-stack/configs/jellyfin:/config
      - /mnt/fastssd/jellyfin/cache:/cache
    restart: unless-stopped
    extra_hosts:
      - host.docker.internal:host-gateway
    runtime: nvidia
    deploy:
      resources:
        reservations:
          devices:
            - capabilities:
                - gpu

To test if your server is properly configured to run docker containers with your NVidia GPU you can run the following command:

sudo docker run --rm --runtime=nvidia --gpus all ubuntu nvidia-smi

This will open NVidia-SMI as a docker container. If you see a table there with information about your GPU and the power consumption etc. you’re good to go! If not, please look up how to configure your docker host for NVidia GPUs.

2.2 Jellyseer

services:
  jellyseerr:
    image: fallenbagel/jellyseerr:latest
    container_name: jellyseerr
    environment:
      - LOG_LEVEL=debug
      - TZ=Europe/Berlin #Changethis to your timezone!
    ports:
      - 5055:5055
    volumes:
      - /mnt/arr-stack/configs/jellyseerr:/app/config
    restart: unless-stopped

3. Configuring each service

After we’ve started our services it’s time to open the webui’s and configure them.

3.1 Jellyfin

  1. Open the webui on http://ipofyourserver:8096
  2. Select Language
  3. Set up username + password
  4. Add a media library
    1. Under content type select the one to start with; Lets assume “Shows”
    2. Give it a “Display Name”
    3. Under Folder click on the + Icon
    4. Browse to /mnt/arr-stack/media/shows and click OK
  5. Add any other Library in the same manner; Movies, Music

3.1.1 Hardware Acceleration

Obviously this part only for if you set jellyfin up with your GPU.

  1. Click the Hamburger Menu in the left upper corner
  2. Go to Settings
  3. Under Administration, click on Dashboard
  4. Click on Playback
  5. Click on Transcoding
  6. Under Hardware Acceleration select “NVidia NVENC”
  7. Select all checkboxes on the different codecs
  8. Enable tone-mapping and keep the algorithm

This is the fundamental basic setup. Jellyfin has an extensive Plugin ecosystem for all different things to do. Most of them make it much more easy to get correct metadata for different types of content.

3.2 Jellyseer

  1. Open the webui http://ipofyourserver:5055
  2. Click on “Use your Jellyfin Account”
  3. Enter your jellyfin URL which is http://ipofyourserver:8096
  4. Enter the credentials you set up in your jellyfin instance
  5. Sign In
  6. Click on “Sync Libraries”
  7. Check your libraries
  8. Click on Start Scan
  9. Click on Continue
  10. Click on Add Radarr Server
  11. Check “Default Server”
  12. Name it “Radarr”
  13. Enter the IP of your Radarr server
  14. Enter your Radarr API Key (you find it in Radarr -> Settings -> General)
  15. Click on Test at the bottom
  16. Select your Quality Profile that you want as a default
  17. Select your Radarr Root Folder
  18. Click on Add Server
  19. Rinse and Repeat with adjusted values for Sonarr

Now if you find a movie or show you can click on “Requst” and see that it will automatically push that request to either radarr or sonarr and they will start looking and downloading for it.


4. You’re done

Now you also have a fully functional and possibly hardware accelerated Jellyfin instance that can do movies, shows and music.

One more interesting addition to this stack would be Jellystats

If you’re looking for something for adult content that goes hand in hand with Whisparr, look into stash and additionally xbvr if you’re looking for a VR library and player.

  • witx@lemmy.sdf.org
    link
    fedilink
    English
    arrow-up
    2
    ·
    2 months ago

    I’ll use this topic to ask a question about jellyseer if you don’t mind.

    I have jellyfin, jellyseer and arr stack for my Linux ISOs. The issue is when one someone requests an ISO from jellyseer it never is the best choice in terms of peers. I can check this by doing interactive search on one of arr and seeing there was a better choice for the quality I setup. Perhaps I have some misconfiguration?

    • RandomLegend [He/Him]@lemmy.dbzer0.comOPM
      link
      fedilink
      English
      arrow-up
      5
      ·
      2 months ago

      Sounds like it, because jellyseer itself does NOT search for anything. All it does is tell sonarr/radarr that something was added and they should start the search.

      However you also have to select your sonarr/radarr profile inside Jellyseer. So if you maybe have an old profile that is set to default on jellyseer it might be it.

      • witx@lemmy.sdf.org
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        2 months ago

        Sorry I’m not very eloquent and failed to explain myself:

        What I see is that the requested “versions” don’t match when the request is made through jellyseer vs when made directly from one of the Arr.

        I first noticed this when requesting through jellyseer and I’d see a file with very few peers. Then I’d do an interactive search in the respective Arr (by hand) and there were much better candidates

        I’ll recheck but I think I have updated profiles

        • RandomLegend [He/Him]@lemmy.dbzer0.comOPM
          link
          fedilink
          English
          arrow-up
          1
          ·
          2 months ago

          I’d make a little test run if i were you.

          Add something via Jellyseer, take a note on what exact release it grabbed and stop the download and remove the whole job. Then add it via *arr again but don’t do a interactive search. Let it do it’s thing and take a note what exact release it grabs. If it takes the same release it did when using jellyseer, then you know somethings not good with your *arr profile here. Because jellyseer basically just prompts *arr to automatically search for something.

          Interactive search is a manual process. If *arr’s automatical search result is bad, there are ways to finetune it. Though i don’t use *arr with torrents so i don’t know exactly where / how.