Network Quality Go Server

Welcome!

The Network Quality Go Server project originally started out as part of the Network Quality Server project. For ease of integration with other projects, it was decided this project would be better served by being it's own repo/project.

Contributing

Please review how to contribute if you would like to submit a pull request.

Asking Questions and Discussing Ideas

If you have any questions you’d like to ask publicly, or ideas you’d like to discuss, please raise a GitHub issue.

Project Maintenance

Project maintenance involves, but is not limited to, adding clarity to incoming issues and reviewing pull requests. Project maintainers can approve and merge pull requests. Reviewing a pull request involves judging that a proposed contribution follows the project’s guidelines, as described by the guide to contributing.

Project maintainers are expected to always follow the project’s Code of Conduct, and help to model it for others.

Project Governance

Although we expect this to happen very infrequently, we reserve the right to make changes, including changes to the configuration format and scope, to the project at any time.

Building (requires Go 1.19+)

make

or

go install ./...

Running

Usage:

Usage of ./networkqualityd:
  -announce
        announce this server using DNS-SD
  -cert-file string
        cert to use
  -config-name string
        domain to generate config for (default "networkquality.example.com")
  -context-path string
        context-path if behind a reverse-proxy
  -create-cert
        generate self-signed certs
  -debug
        enable debug mode
  -enable-cors
        enable CORS headers
  -enable-h2c
        enable h2c (non-TLS http/2 prior knowledge) mode
  -enable-http2
        enable HTTP/2 (default true)
  -enable-http3
        enable HTTP/3
  -enable-l4s
        Enable L4S using the default congestion control algorithm, prague.
  -enable-l4s-algorithm string
        Enable L4S using the specified congestion control algorithm
  -insecure-public-port int
        The port to listen on for HTTP measurement accesses
  -key-file string
        key to use
  -listen-addr string
        address to bind to (default "localhost")
  -public-name string
        host to generate config for (same as -config-name if not specified)
  -public-port int
        The port to listen on for HTTPS/H2C/HTTP3 measurement accesses (default 4043)
  -socket-send-buffer-size uint
        The size of the socket send buffer via TCP_NOTSENT_LOWAT. Zero/unset means to leave unset
  -tos string
        set TOS for listening socket (default "0")
  -version
        Show version

Example run:

./networkqualityd --create-cert --public-name networkquality.example.com
2023/05/01 10:38:56 Network Quality URL: https://networkquality.example.com:4043/.well-known/nq
2023/05/01 10:38:56 Enabling H2 on "localhost:4043"

Running Apple's client against server:

networkQuality -C https://networkquality.example.com:4043/.well-known/nq -k
==== SUMMARY ====
Uplink capacity: 1.649 Gbps
Downlink capacity: 4.933 Gbps
Responsiveness: Medium (606 RPM)
Idle Latency: 3.917 milliseconds

Running the goresponsiveness client against server:

From the gorespsonsiveness checkout

go run networkQuality.go --url https://networkquality.example.com:4043/.well-known/nq --insecure-skip-verify
05-01-2023 17:42:14 UTC Go Responsiveness to networkquality.example.com:4043...
RPM:  1197 (P90)
RPM:  1907 (Double-Sided 10% Trimmed Mean)
Download: 6038.648 Mbps (754.831 MBps), using 9 parallel connections.
Upload:   1561.561 Mbps (195.195 MBps), using 9 parallel connections.

Docker

The server can be run in a docker container. The Dockerfile in this repository will generate a container that can run the server. To build the container, simply execute

docker build -t rpmserver .

The command will generate a container image named rpmserver.

In order to run the resulting container, you will have to either accept some default values or provide configuration. The server requires access to a public/private key for its SSL connections, and the Dockerfile does not specify that they be copied in to the image. In other words, you will have to configure a shared volume between the executing container and the host, where that volume contains the key files.

The container executing the RPM server will also need to have a port map established. You will have to publish the port on which the server in the container is listening to the host.

Assuming that you use the default values specified in the Dockerfile, you can run the container using

docker run --env-file docker_config.env  -v $(pwd)/live:/live -p 4043:4043 -p rpmserver

where there exists a directory $(pwd)/live that contains two files named fullchain.pem and privkey.pem that hold the public and private keys for the SSL connections, respectively.

You can use environment variables to configure any of the networkqualityd command-line options.

Command-line option name	Environment variable name
-cert-file	cert_file
-key-file	key_file
-public-port	public_port
-config-name	config_name
-listen-addr	listen_addr
-public-name	public_name
-debug	see below

If you want to configure whether the server runs in debug mode, simply set the debug environment variable to -debug. If you enable debugging, you will also need to create a map between a port on the host and port 9090 on the container (e.g., -p 9090:9090).

There is docker_config.env in this directory that you can use to make passing those configuration options to the container easier. To use this file, add the --env-file docker_config.env arguments to the docker run command.