Avalanche-Node

Deployment for an Avalanche node using Kubernetes.

For the purpose of this exercise, following tools will be used

• docker
• minikube
• golang (>=1.17.9 except v1.18.x)
• helm3

Startup minikube

For this exercise, minikube will be used to run kubernetes cluster locally on ubuntu 20.04 LTS

minikube start --cpus 8 --memory 10g

Build Avalanche Node docker image

Using the scripts provided from Avalanchego github to build docker image from source https://github.com/ava-labs/avalanchego#docker-install

$ git clone https://github.com/ava-labs/avalanchego.git

$ ./scripts/build_image.sh

$ docker image ls 

Am image with name and tag avalanchego:master will be created

Deploy Avalanche Node

Upload docker image builded from source to minikube

minikube image load avalanchego:master

Deployment k8s resources - deployment, PV, PVC and service

kubectl apply -f .

Install Grafana/Prometheus onto k8s

To monitor this instances of Avalanche node, we use Grafana/Premetheus to gather metrics and visualize data.

First, we'll need to install Prometheus Operator via the community helm chart

$ helm repo add prometheus-community https://prometheus-community.github.io/helm-charts

$ helm repo update

$ helm install monitoring prometheus-community/kube-prometheus-stack

Login into Grafana

# expose grafana UI http endpoint 
$ kubectl port-forward svc/monitoring-grafana 8080:80

# Default Username is admin
# Get password 
$ kubectl get secret  monitoring-grafana -o json | jq -r .'["data"]["admin-password"]' | base64 --decode

Once logged into Grafana UI, we can upload and use pre-made Dashboards

Source Reference: https://docs.avax.network/nodes/maintain/setting-up-node-monitoring