MD Chat

Auth backend

• MD Chat relies on Logto as the authentication backend, which is a powerful open-source identity and access management solution.
• You may choose to self-host Logto or use their cloud service.

Note

MD Chat has only been tested to work with a self-hosted Logto instance, but Logto Cloud should work just as well.

• Once you have Logto set up, create a Traditional Web app with the following settings:

  • Name: A name for your app, e.g. “MD Chat”
  • Redirect URI: http://(your-app-url)/callback

• Keep the Client ID and Client Secret handy, as you will need them to configure MD Chat.

Deployment

Caution

Due to a reliance on a traditional Postgres database, MD Chat is designed to be deployed as a Docker container and may not work on serverless platforms like Vercel or Cloudflare Workers.

To start, you will need the following:

• A Logto instance and app credentials as described above.
• A Postgres database
• A Redis instance
• An S3-compatible object storage (e.g. AWS S3, Tigris, Cloudflare R2, etc.) for storing uploaded files.

The Docker image is available on GitHub, and can be pulled with the following command:

docker pull ghcr.io/mikndotdev/chat:latest

Setup the environment variables in your Docker container. A sample .env file might look like this:

DATABASE_URL="postgres://postgres:password@1.2.3.4:5402/postgres"
REDIS_URL="redis://:username@5.6.7.8:5409/0"
LOGTO_APP_ID="your-logto-app-id"
LOGTO_APP_SECRET="your-logto-app-secret"
LOGTO_ENDPOINT="https://(your-logto-instance).logto.io"
LOGTO_COOKIE_SECRET="a-random-secret-string"
S3_ENDPOINT="https://s3.example.com"
S3_ACCESS_KEY_ID="your-s3-access-key-id"
S3_SECRET_ACCESS_KEY="your-s3-secret-access-key"
S3_BUCKET_NAME="your-s3-bucket-name"
S3_REGION="your-s3-region"
S3_UPLOAD_DIR="any-directory-you-want"
S3_PUBLIC_URL="https://s3.example.com/your-s3-bucket-name/"

Note

Make sure to set the S3_PUBLIC_URL to the public URL of your S3 bucket, as this is used to serve uploaded files. If you have a seperate CDN behind your bucket, set that as the public URL.

You can then run the container with the following command:

docker run -d \
  --name md-chat \
  -p 3000:80 \
  --env-file .env \
  ghcr.io/mikndotdev/chat:latest

• This will start the MD Chat server on port 3000, and you can access it at http://localhost:3000.
• Make sure to adjust the port and other settings as needed for your deployment environment.

Tip

I’m a huge fan of Coolify, which is an amazing piece of free software makes deploying, managing and proxiying Docker containers to your infrstructure a breeze. It’s what the MD Chat demo and MikanDev’s Logto instance is running on, and I highly recommend it for self-hosting!

Docker Compose

Caution

While this setup is great to quickly get started, for production deployments it is recommended to manually set up each service and configure them according to your needs.

• An example Docker Compose file is available in the repository, which has all the necessary services configured except for Logto and S3.
• You can use this file to quickly set up a local development environment with MD Chat, and the databases.