mc-router/README.md

Routes Minecraft client connections to backend servers based upon the requested server address.

Usage

  -api-binding host:port
    	The host:port bound for servicing API requests (env API_BINDING)
  -auto-scale-up
      Increase Kubernetes StatefulSet Replicas (only) from 0 to 1 on respective backend servers when accessed (env AUTO_SCALE_UP)
  -connection-rate-limit int
    	Max number of connections to allow per second (env CONNECTION_RATE_LIMIT) (default 1)
  -cpu-profile string
    	Enables CPU profiling and writes to given path (env CPU_PROFILE)
  -debug
    	Enable debug logs (env DEBUG)
  -in-kube-cluster
    	Use in-cluster Kubernetes config (env IN_KUBE_CLUSTER)
  -kube-config string
    	The path to a Kubernetes configuration file (env KUBE_CONFIG)
  -mapping string
    	Comma-separated mappings of externalHostname=host:port (env MAPPING)
  -metrics-backend string
    	Backend to use for metrics exposure/publishing: discard,expvar,influxdb (env METRICS_BACKEND) (default "discard")
  -metrics-backend-config-influxdb-addr string
    	 (env METRICS_BACKEND_CONFIG_INFLUXDB_ADDR)
  -metrics-backend-config-influxdb-database string
    	 (env METRICS_BACKEND_CONFIG_INFLUXDB_DATABASE)
  -metrics-backend-config-influxdb-interval duration
    	 (env METRICS_BACKEND_CONFIG_INFLUXDB_INTERVAL) (default 1m0s)
  -metrics-backend-config-influxdb-password string
    	 (env METRICS_BACKEND_CONFIG_INFLUXDB_PASSWORD)
  -metrics-backend-config-influxdb-retention-policy string
    	 (env METRICS_BACKEND_CONFIG_INFLUXDB_RETENTION_POLICY)
  -metrics-backend-config-influxdb-tags value
    	any extra tags to be included with all reported metrics (env METRICS_BACKEND_CONFIG_INFLUXDB_TAGS)
  -metrics-backend-config-influxdb-username string
    	 (env METRICS_BACKEND_CONFIG_INFLUXDB_USERNAME)
  -port port
    	The port bound to listen for Minecraft client connections (env PORT) (default 25565)
  -version
    	Output version and exit (env VERSION)

REST API

• GET /routes (with Accept: application/json)

  Retrieves the currently configured routes

• POST /routes (with Content-Type: application/json)

  Registers a route given a JSON body structured like:

  {
    "serverAddress": "CLIENT REQUESTED SERVER ADDRESS",
    "backend": "HOST:PORT"
  }
  

• POST /defaultRoute (with Content-Type: application/json)

  Registers a default route to the given backend. JSON body is structured as:

  {
    "backend": "HOST:PORT"
  }
  

• DELETE /routes/{serverAddress}

  Deletes an existing route for the given serverAddress

Docker Multi-Architecture Image

The multi-architecture image published at Docker Hub supports amd64, arm64, and arm32v6 (i.e. RaspberryPi).

Docker Compose Usage

The following diagram shows how the example docker-compose.yml configures two Minecraft server services named vanilla and forge, which also become the internal network aliases. Notice those services don't need their ports exposed since the internal networking allows for the inter-container access.

The router service is only one of the services that needs to exposed on the external network. The --mapping declares how the hostname users will enter into their Minecraft client will map to the internal services.

To test out this example, I added these two entries to my "hosts" file:

127.0.0.1 vanilla.example.com
127.0.0.1 forge.example.com

Kubernetes Usage

Using Kubernetes Service auto-discovery

When running mc-router as a Kubernetes Pod and you pass the --in-kube-cluster command-line argument, then it will automatically watch for any services annotated with

• mc-router.itzg.me/externalServerName : The value of the annotation will be registered as the external hostname Minecraft clients would used to connect to the routed service. The service's clusterIP and target port are used as the routed backend. You can use more hostnames by splitting them with comma.
• mc-router.itzg.me/defaultServer : The service's clusterIP and target port are used as the default if no other externalServiceName annotations applies.

For example, start mc-router's container spec with

image: itzg/mc-router
name: mc-router
args: ["--in-kube-cluster"]

and configure the backend minecraft server's service with the annotation:

apiVersion: v1
kind: Service
metadata:
  name: mc-forge
  annotations:
    "mc-router.itzg.me/externalServerName": "external.host.name"

you can use multiple host names:

apiVersion: v1
kind: Service
metadata:
  name: mc-forge
  annotations:
    "mc-router.itzg.me/externalServerName": "external.host.name,other.host.name"

Example Kubernetes deployment

This example deployment

• Declares an mc-router service that exposes a node port 25565
• Declares a service account with access to watch and list services
• Declares --in-kube-cluster in the mc-router container arguments
• Two "backend" Minecraft servers are declared each with an "mc-router.itzg.me/externalServerName" annotation that declares their external server name(s)

kubectl apply -f https://raw.githubusercontent.com/itzg/mc-router/master/docs/k8s-example-auto.yaml

Notes

• This deployment assumes two persistent volume claims: mc-stable and mc-snapshot
• I extended the allowed node port range by adding --service-node-port-range=25000-32767 to /etc/kubernetes/manifests/kube-apiserver.yaml

Auto Scale Up

The -auto-scale-up flag argument makes the router "wake up" any stopped backend servers, by changing replicas: 0 to replicas: 1.

This requires using kind: StatefulSet instead of kind: Service for the Minecraft backend servers.

It also requires the ClusterRole to permit get + update for statefulsets & statefulsets/scale, e.g. like this (or some equivalent more fine-grained one to only watch/list services+statefulsets, and only get+update scale):

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: services-watcher
rules:
- apiGroups: [""]
  resources: ["services"]
  verbs: ["watch","list"]
- apiGroups: ["apps"]
  resources: ["statefulsets", "statefulsets/scale"]
  verbs: ["watch","list","get","update"]

Development

Building locally with Docker

docker build -t mc-router .

Build locally without Docker

After installing Go and doing a go mod download to install all required prerequisites, just like the Dockerfile does, you can:

make test # go test -v ./...
go build ./cmd/mc-router/

Skaffold

For "in-cluster development" it's convenient to use https://skaffold.dev. Any changes to Go source code will trigger a go build, new container image pushed to registry with a new tag, and refresh in Kubernetes with the image tag used in the deployment transparently updated to the new tag and thus new pod created pulling new images, as configured by skaffold.yaml:

skaffold dev

When using Google Cloud (GCP), first create a Docker Artifact Registry, then add the Artifact Registry Reader Role to the Compute Engine default service account of your GKE clusterService Account (to avoid error like "container mc-router is waiting to start: ...-docker.pkg.dev/... can't be pulled"), then use e.g. gcloud auth configure-docker europe-docker.pkg.dev or equivalent one time (to create a ~/.docker/config.json), and then use e.g. --default-repo=europe-docker.pkg.dev/YOUR-PROJECT/YOUR-ARTIFACT-REGISTRY option for skaffold dev.

Performing snapshot release with Docker

docker run -it --rm \
  -v ${PWD}:/build -w /build \
  -v /var/run/docker.sock:/var/run/docker.sock \
  goreleaser/goreleaser \
  release --snapshot --rm-dist

Related Projects

• https://github.com/haveachin/infrared