/knightcrawler

A selfhosted Stremio addon

Primary LanguageJavaScriptApache License 2.0Apache-2.0

isolated logo

Knight Crawler

A self-hosted Stremio addon for streaming torrents via a debrid service.

Join our Discord !

Contents

Important

The latest change renames the project and requires a small migration.

Overview

Stremio is a media player. On it's own it will not allow you to watch anything. This addon at it's core does the following:

  1. It will search the internet and collect information about movies and tv show torrents, then store it in a database.
  2. It will then allow you to click on the movie or tv show you desire in Stremio and play it with no further effort.

Using

The project is shipped as an all-in-one solution. The initial configuration is designed for hosting only on your local network. If you want it to be accessible from outside of your local network, please see Configure external access.

Download Docker and Docker Compose v2

Download and install Docker Compose, bundled with Docker Desktop or, if using Linux, you can install Docker Engine and the Docker Compose Plugin.

Environment Setup

Before running the project, you need to set up the environment variables. Edit the values in stack.env:

cd deployment/docker
code stack.env

Then set any of the values you wouldd like to customize.

Optional Configuration Changes

Warning

These values should be tested and tuned for your specific machine.

By default, Knight Crawler is configured to be relatively conservative in its resource usage. If running on a decent machine (16GB RAM, i5+ or equivalent), you can increase some settings to increase consumer throughput. This is especially helpful if you have a large backlog from importing databases.

DebridMediaManager setup (optional)

There are some optional steps you should take to maximise the number of movies/tv shows we can find.

We can search DebridMediaManager hash lists which are hosted on GitHub. This allows us to add hundreds of thousands of movies and tv shows, but it requires a Personal Access Token to be generated. The software only needs read access and only for public repositories. To generate one, please follow these steps:

  1. Navigate to GitHub settings -> Developer Settings -> Personal access tokens -> Fine-grained tokens (click here for a direct link)
  2. Press Generate new token
  3. Fill out the form (example data below):
     Token name:
         KnightCrawler
     Expiration:
         90 days
     Description:
         <blank>
     Repository access
         (checked) Public Repositories (read-only)
    
  4. Click Generate token
  5. Take the new token and add it to the bottom of the stack.env file
    GITHUB_PAT=<YOUR TOKEN HERE>
    

Configure external access

Please choose which applies to you:

I have a public IP address and can open ports

You can use either a paid domain your-domain.com or a free reverse dns service like DuckDNS (you can automate the update of your IP address).

Before continuing you need to open up port 80 and 443 in your firewall and configure any port forwarding as necessary. You should not do this unless you understand the security implications. Please note that Knightcrawler and its contributors cannot be held responsible for any damage or loss of data from exposing your service publicly.

You may find it safer to use a tunnel/vpn, but this will require the use of a paid domain or will not be accessible without being connected to your vpn.

I will be using a tunnel/vpn (CGNAT, don't want to open ports, etc...)

For this you can use a VPN like Tailscale which has its own ways of issuing SSL certs, or you can use a tunnel like Cloudflare.

To use a Cloudflare tunnel you will need a domain name.

There's a sample compose for a Cloudflare tunnel here.

If you are going to go this route, you will want to connect caddy to the cloudflare-tunnel network. It's all in Caddy's docker-compose.yaml you will just need to uncomment it.

Next steps

Regardless of what method you choose, you will need to connect Knight Crawler to Caddy. We only need to expose the addon, the rest of the services can remain internal.

In our primary docker-compose.yaml we will add the Caddy network:

networks:
  knightcrawler-network:
    driver: bridge
    name: knightcrawler-network

  caddy:
    name: caddy
    external: true

Remove or comment out the port for the addon, and connect it to Caddy:

addon:
  <<: *knightcrawler-app
  env_file:
    - stack.env
  hostname: knightcrawler-addon
  image: gabisonfire/knightcrawler-addon:latest
  labels:
    logging: "promtail"
  networks:
    - knightcrawler-network
    - caddy         # <~~~~~~~ These lines
  # ports:          # <~~~~~~~ have been
  #   - "7000:7000" # <~~~~~~~ changed

If you are using a Cloudflare tunnel, start it before Caddy.

Caddy can be started with:

cd deployment/docker/optional_reverse_proxy
docker compose up -d

It should be started before Knight Crawler.

Run the project

To start the project use the following commands:

cd deployment/docker
docker compose up -d

It will take a while to find and add the torrents to the database. During initial testing, in one hour it's estimated that around 200,000 torrents were located and added to the queue to be processed. The processing takes longer, unfortunately and you may not find the movie/show you want for a while. For best results, you should leave everything running for a few hours.

To add the addon to Stremio, open a web browser and navigate to: http://127.0.0.1:7000 or knightcrawler.your-domain.com if you are using Caddy.

Monitoring with Grafana and Prometheus (Optional)

To enhance your monitoring capabilities, you can use Grafana and Prometheus in addition to RabbitMQ's built-in management interface. This allows you to visualize and analyze RabbitMQ metrics with more flexibility. With postgres-exporter service, you can also monitor Postgres metrics.

Accessing RabbitMQ Management

You can still monitor RabbitMQ by accessing its management interface at http://127.0.0.1:15672/. Use the provided credentials to log in and explore RabbitMQ's monitoring features (the default username and password are guest).

Using Grafana and Prometheus

Here's how to set up and use Grafana and Prometheus for monitoring RabbitMQ:

  1. Start Grafana and Prometheus: Run the following command to start both Grafana and Prometheus:

    cd deployment/docker/optional_metrics
    docker compose up -d
  2. Import Grafana Dashboard: Import the RabbitMQ monitoring dashboard into Grafana:

    The Prometheus data source is already configured in Grafana, you just have to select it when importing the dashboard.

Now, you can use these dashboards to monitor RabbitMQ and Postgres metrics.

Note

If you encounter issues with missing or unavailable data in Grafana, please ensure on Prometheus's target page that the RabbitMQ target is up and running.

Importing external dumps

If you stumble upon some data, perhaps a database dump from rarbg, which is almost certainly not available on Real Debrid, and wished to utilise it, you might be interested in the following:

Importing data into PostgreSQL

Using pgloader we can import other databases into Knight Crawler.

For example, if you possessed a sqlite database named rarbg_db.sqlite it's possible to then import this.

Using pgloader via docker


Make sure the file rarbg_db.sqlite is on the same server as PostgreSQL. Place it in any temp directory, it doesn't matter, we will delete the these files afterwards to free up storage.

For this example I'm placing my files in ~/tmp/load

So after uploading the file, I have ~/tmp/load/rarbg_db.sqlite. We need to create the file ~/tmp/load/db.load which should contain the following:

load database
     from sqlite:///data/rarbg_db.sqlite
     into postgresql://postgres:postgres@postgres:5432/knightcrawler

with include drop, create tables, create indexes, reset sequences

  set work_mem to '16MB', maintenance_work_mem to '512 MB';

Note

If you have changed the default password for PostgreSQL (RECOMMENDED) please change it above accordingly

postgresql://postgres:<password here>@postgres:5432/knightcrawler

Then run the following docker run command to import the database:

docker run --rm -it --network=knightcrawler-network -v "$(pwd)":/data dimitri/pgloader:latest pgloader /data/db.load

Important

The above line does not work on ARM devices (Raspberry Pi, etc). The image does not contain a build that supports ARM.

I found jahan9/pgloader which supports ARM64 but I offer no guarantees that this is a safe image.

USE RESPONSIBLY

The command to utilise this image would be:

docker run --rm -it --network=knightcrawler-network -v "$(pwd)":/data jahan9/pgloader:latest pgloader /data/db.load

Using native installation of pgloader


Similarly to above, make sure the file rarbg_db.sqlite is on the same server as PostgreSQL. Place it in any temp directory, it doesn't matter, we will delete the these files afterwards to free up storage.

For this example I'm placing my files in ~/tmp/load

So after uploading the file, I have ~/tmp/load/rarbg_db.sqlite. We need to create the file ~/tmp/load/db.load which should contain the following:

load database
     from sqlite://~/tmp/load/rarbg_db.sqlite
     into postgresql://postgres:postgres@<docker-ip>/knightcrawler

with include drop, create tables, create indexes, reset sequences

  set work_mem to '16MB', maintenance_work_mem to '512 MB';

Note

If you have changed the default password for PostgreSQL (RECOMMENDED) please change it above accordingly

postgresql://postgres:<password here>@<docker-ip>/knightcrawler

Tip

Your docker-ip can be found using the following command:

docker network inspect knightcrawler-network | grep knightcrawler-postgres -A 4

Then run pgloader db.load.

Process the data we have imported

The data we have imported is not usable immediately. It is loaded into a new table called items. We need to move this data into the ingested_torrents table to be processed. The producer and at least one consumer need to be running to process this data.

We are going to concatenate all of the different movie categories into one e.g. movies_uhd, movies_hd, movies_sd -> movies. This will give us a lot more data to work with.

The following command will process the tv shows:

docker exec -it knightcrawler-postgres-1 psql -U postgres -d knightcrawler -c "
INSERT INTO ingested_torrents (name, source, category, info_hash, size, seeders, leechers, imdb, processed, \"createdAt\", \"updatedAt\")
SELECT title, 'RARBG', 'tv', hash, size, NULL, NULL, imdb, false, current_timestamp, current_timestamp
FROM items where cat LIKE 'tv%%' ON CONFLICT DO NOTHING;"

This will do the movies:

docker exec -it knightcrawler-postgres-1 psql -U postgres -d knightcrawler -c "
INSERT INTO ingested_torrents (name, source, category, info_hash, size, seeders, leechers, imdb, processed, \"createdAt\", \"updatedAt\")
SELECT title, 'RARBG', 'movies', hash, size, NULL, NULL, imdb, false, current_timestamp, current_timestamp
FROM items where cat LIKE 'movies%%' ON CONFLICT DO NOTHING;"

Both should return a response that looks like:

INSERT 0 669475

As soon as both commands have been run, we can immediately drop the items table. The data inside has now been processed, and will just take up space.

docker exec -it knightcrawler-postgres-1 psql -U postgres -d knightcrawler -c "drop table items";

I imported the data without the LIKE 'movies%%' queries!

If you've already imported the database using the old instructions, you will have imported a ton of new movies and tv shows, but the new queries will give you a lot more!

Don't worry, we can still correct this. We can retroactively update the categories by running the following commands

# Update movie categories
docker exec -it knightcrawler-postgres-1 psql -d knightcrawler -c "UPDATE ingested_torrents  SET category='movies', processed='f' WHERE category LIKE 'movies_%';"

# Update TV categories
docker exec -it knightcrawler-postgres-1 psql -d knightcrawler -c "UPDATE ingested_torrents  SET category='tv', processed='f' WHERE category LIKE 'tv_%';"

Selfhostio to KnightCrawler Migration

With the renaming of the project, you will have to change your database name in order to keep your existing data.

With your existing stack still running, run:

docker exec -it torrentio-selfhostio-postgres-1 psql -c "
SELECT pg_terminate_backend(pid) FROM pg_stat_activity
WHERE pid <> pg_backend_pid() AND datname = 'selfhostio';
ALTER DATABASE selfhostio RENAME TO knightcrawler;"

Make sure your postgres container is named torrentio-selfhostio-postgres-1, otherwise, adjust accordingly.

This command should return: ALTER DATABASE. This means your database is now renamed. You can now pull the new changes if you haven't already and run docker compose up -d.

To-do

  • Add a troubleshooting section