GoodWatch 开源影视推荐项目：基于评分的个性化推荐与流媒体聚合

GoodWatch Online

GoodWatch is alive and running. Check it out: https://goodwatch.app/

Roadmap

The living roadmap organizes the backlog and guides evolving priorities, using periodic research and results to reassess the current focus.

GoodWatch Community

Join our Discord to be part of the wonderful community and learn more about the project: https://discord.gg/TVAcrfQzcA

What's unique about GoodWatch?

Personalized Recommendations based on your ratings and watched movies & shows
Fingerprint: Classify movies and TV shows by hundreds of meaningful categories (e.g. Emotions, Plot, Subgenres, Visual Style, Cultural Context, etc.)
See availability of streaming services for each title (Netflix, Hulu, Amazon Prime, ...)
See all ratings at one place (IMDb, Metacritic, Rotten Tomatoes)

Repository Structure

Directory	Description
`goodwatch-benchmark/`	Load tests and Lighthouse before-and-after comparisons
`goodwatch-webapp/`	Remix web application (frontend + API)
`goodwatch-hq/`	Admin tools and internal services
`goodwatch-remote/`	Ansible deployment scripts and Windmill Workers
`goodwatch-flows/`	Windmill data pipelines and ETL workflows
`goodwatch-qdrant/`	Qdrant vector database for similarity search and recommendations
`goodwatch-crate/`	CrateDB for webapp access
`goodwatch-mongo/`	MongoDB for data pipeline storage
`goodwatch-cache/`	Redis cache cluster configuration
`goodwatch-metrics/`	Prometheus metrics collection
`goodwatch-monitoring/`	Grafana dashboards and alerting
`goodwatch-proxy/`	Caddy reverse proxy configuration

Contributing

We welcome contributions! Please feel free to submit a Pull Request.

Run Locally

Documentation Status

The following guide is outdated. If you want to contribute to the project, please join the Discord Community.

Getting started

Running the web app:

Clone this repository
Register for a TMDB API Key
Copy .env.example to .env from /goodwatch-webapp and fill out the required secrets

You are now ready to run it locally:

cd goodwatch-webapp
npm install
npm run dev

Server Guide

Running the whole infrastructure to run GoodWatch is more complex. The guide below is far from complete at the moment.

Cache Cluster

Three Valkey 8 nodes in cluster mode, from the Compose files in goodwatch-cache/. See docs/private-redis.md for the configuration and the procedures.

Data Pipeline DB Cluster

https://medium.com/workleap/the-only-local-mongodb-replica-set-with-docker-compose-guide-youll-ever-need-2f0b74dd8384

openssl rand -base64 756 > mongodb-keyfile

https://github.com/csuka/ansible_role_mongodb_ubuntu

ansible-galaxy collection install community.mongodb
ansible-galaxy install csuka.mongodb_ubuntu
ansible-playbook -i inventory.ini playbook.yml

Windmill Server & Workers

TODO

Adding a new Server

Install Packages

apt-get install make fzf

Install Docker

https://docs.docker.com/engine/install/ubuntu/#install-using-the-repository

Git Repository

git clone https://github.com/alp82/goodwatch-monorepo.git

Grafana Exporter

Windmill Worker

Backup

sshfs

apt-get update && apt-get install sshfs

fstab

u123456-subX@u123456-subX.your-storagebox.de:/home /mnt/backup-xxx fuse.sshfs _netdev,port=23,IdentityFile=/root/.ssh/id_rsa,allow_other,reconnect,ServerAliveInterval=15,ServerAliveCountMax=3 0 0

curl -LsSf https://astral.sh/uv/install.sh | sh

Troubleshooting

Not enough disc space

docker image prune -a
docker builder prune

Relocate docker root directory

https://www.ibm.com/docs/en/z-logdata-analytics/5.1.0?topic=compose-relocating-docker-root-directory