mineflayer-k8s

A Mineflayer example bot runnable as a baremetal process, as a Docker container or as a Kubernetes Deployment.

---

---

Detect toxic players using TensorflowJS

If you're lucky mineflayer-k8s(s) are running on 91.121.67.56:30018 Minecraft server running in a Kubernetes cluster running on a Raspberry PI.

Viewer

You can interactively see what your bot is doing using prismarine-viewer in the web.
See example here You can disable this feature easily by adding viewer to the disabled plugins (or any plugins in the plugins directory) in the configuration:

"disabledPlugins": ["viewer"]

Node dependencies

• mineflayer, mineflayer-cmd, mineflayer-pathfinder, mineflayer-pvp, prismarine-viewer (Minecraft bots)
• @tensorflow/tfjs-node, @tensorflow/tfjs, @tensorflow-models/toxicity (deep learning)
• winston (logging)
• yargs (arguments)

Usage

If you want to use Kubernetes and know it plus Helm, just jump directly to the values.

louis@louis-pc:~/Documents/mineflayer-k8s$ make help
usage: make [target] ...

Variables:
TAG: Docker tag, default "TAG ?= 0.1.4"
CONFIG: path to application config file, default "CONFIG ?= setting.json"
SERVER_CONFIG: path to minecraft server config file, default "SERVER_CONFIG ?= server-values.yaml"

Targets:
help                   Display this help
build_push             Build and push on DockerHub multiarch Docker image (need Docker buildx)
deploy                 Deploy as a baremetal process
deploy_docker          Deploy as a Docker container
deploy_k8s             Deploy as a Kubernetes Deployment
deploy_k8s_server      Deploy a github.comitzg/minecraft-server-charts on Kubernetes

Run as nodejs process

Dependencies

• make
• nodejs

Default settings

Using default settings

make deploy

With custom settings

CONFIG=./setting.json make deploy

Run as Docker container

Dependencies

• make
• docker
• Supported OS/ARCH see tags, I personally run this on my Raspberry PI 4B

Default settings

Using default settings

make deploy_docker

With custom settings

CONFIG=./setting.json make deploy_docker

Run as Kubernetes Deployment

Dependencies

• make
• A k8s cluster running and configured
• jq
• helm

Default settings

Using default settings

make deploy_k8s

With custom settings

RELEASE=mt1 CONFIG=./setting.json make deploy_k8s

Example Helm output

helm install mk helm -f helm/values-dev.yaml --dry-run --debug

NOTES: mk connecting as foo to 109.210.246.114:30018, server running >version 1.16.4

1. Get logs kubectl logs -f -l app.kubernetes.io/instance=mk
2. Watch your bot playing in the web using prismarine-viewer at 109.210.246.114:30019

Deploy a externally accessible Minecraft server on your k8s cluster

1. Port-forward [port between 30000 and 31000 (k8s allowed)] TCP, Kubernetes API (6443) TCP

2. Assign a static IP

3. Use similar config for make deploy_k8s_server:

   minecraftServer:
     eula: true
     version: "1.16.4"
     serviceType: LoadBalancer
     loadBalancerIP: <curl ifconfig.me on your k8s host>
     externalIPs:
       - <curl ifconfig.me on your k8s host>
     nodePort: <my port forwarded port>

Troubleshoot

• Could not resolve host: launchermeta.mojang.com - see issue

  kubectl edit deployment/minecraft-server-minecraft
  dnsPolicy: "None"
  dnsConfig:
    nameservers:
      - 8.8.8.8