alexhaydock/boringnginx

This container builds the latest stable Nginx with the latest BoringSSL code. It was created to aid with the easy deployment of TLS 1.3 services at a time when most Linux distributions were not packaging a version of OpenSSL that could handle it.

This container is built automatically using GitLab CI and supports the x86_64 and aarch64 architectures.

This container builds Nginx with the following modules:

• ngx_brotli
• ngx_headers_more
• ngx_http2_geoip (when built locally with appropriate GeoIP.conf)
• ngx_subs_filter

There are versions of this container which build against:

• BoringSSL (this container)
• LibreSSL
• OpenSSL

Quick test this container locally

Run this container as a quick test (it will listen on http://127.0.0.1 and you will see logs directly in the terminal when connections are made):

docker run --rm -it -p "127.0.0.1:80:80/tcp" registry.gitlab.com/alexhaydock/boringnginx:$(uname -m)

Running with your own config file

Run this container as a daemon with your own config file:

docker run -d -p "80:80/tcp" -p "443:443/tcp" -v /path/to/nginx.conf:/etc/nginx.conf:ro registry.gitlab.com/alexhaydock/boringnginx:$(uname -m)

Build This Container Locally

I cannot distribute the MaxMind GeoIP databases legally with this project, so if you want GeoIP features, you must build this container locally. Sign up for a free account with MaxMind and follow the instructions to generate your own GeoIP.conf and place it in this directory.

Now run:

make geoip

Or you can build the regular container with just:

make build

Running Without Root

You can lock down this container and run without root and dropping all capabilities by using the --user and --cap-drop=ALL arguments.

For this example to work, your config file should instruct Nginx to listen on port 8080 inside the container:

docker run --rm -it -p "80:8080/tcp" --user 6666 --cap-drop=ALL -v /path/to/nginx.conf:/etc/nginx.conf:ro registry.gitlab.com/alexhaydock/boringnginx:$(uname -m)

You will need to make sure that the UID you pick matches the one you have set as the NGINX_ID in the Dockerfile, and that any configs which you mount into the container are owned by this UID (it does not need to exist on the host system).

If you are running rootless like this, you will also want to ensure that the nginx.conf does not try to listen on any ports below 1000 (you can still listen on :80 and :443 externally since the Docker daemon runs as root and can handle this - Nginx does not need to).