Chen188 / eth2-chart

Helm chart to run your own non-staking eth2 node

Amazon EKS is good to run Ethereum node, Helm could be used to simplify the deployment work.

Prerequisites

• Create Amazon EKS cluster with at least one node group
• Instance types used in node group should meet the requirements to run both geth and lighthouse together, m6i.2xlarge (has 8 vCPUs and 32GB memory) is a good point to start with for mainnet.
• Install and config eksctl and kubectl locally, follow this document if needed
• Install Helm, https://helm.sh/docs/intro/install/

Steps to setup Ethereum node

We'll use the default settings for demostration, if you want to use storage class other than gp2, enable ingress controller, or setup monitoring function, check Other Considerations first.

If you decide to continue with gp2 storage class, make sure you've installed the ebs-csi-controller in your EKS cluster, if not, follow this document first.

Clone the code

git clone https://github.com/Chen188/eth2-chart

Create a dedicated k8s namespace

NAMESPACE=testnet
kubectl create namespace $NAMESPACE

Create a JWT secret

JWT secret is required by geth and lighthouse to authenticate the RPC connection.

kubectl create secret generic eth2-jwt-secret --from-literal=jwt-secret=$(openssl rand -hex 32) -n $NAMESPACE

Deploy Ethereum node

# for sepolia testnet
helm install my-eth2 -f charts/values-sepolia.yaml ./charts/ -n $NAMESPACE

# for mainnet
# helm install my-eth2 -f charts/values-mainnet.yaml ./charts/ -n $NAMESPACE

You can check the deployment status with:

kubectl get statefulset -n $NAMESPACE 

NAME                      READY   AGE
my-eth2-geth-lighthouse   1/1     99m

Other Considerations

Storage Class

There're many storage options to storing blockchain data, incluing gp2(default), gp3(current generation of general purpose EBS), FSx for OpenZFS and ephemeral instance storage. Here for demostration, we'll use default gp2 storage class , you can also follow the documents below if you want to try other options:

• gp3: https://aws.amazon.com/cn/blogs/containers/migrating-amazon-eks-clusters-from-gp2-to-gp3-ebs-volumes/

• FSx for OpenZFS: https://github.com/kubernetes-csi/csi-driver-nfs

ALB ingress controller

You can expose geth's HTTP and Websocket endpoint through AWS ALB(Application Load Balancer).

To do so, set services[*].ingress.enabled to true in values-{sepolia,mainnet}.yaml file, and follow the document Application load balancing on Amazon EKS, Installing the AWS Load Balancer Controller add-on to setup alb ingress controller.

After ALB ingress is deployed, you can test the endpoint using command below(replace the host with your ALB DNS name):

curl -X POST \
-H "Content-Type: application/json" \
--data '{"jsonrpc": "2.0", "id": 1, "method": "eth_syncing", "params": []}' \
"http://k8s-eth2-xxxx-7eaf279140-1407294883.us-east-1.elb.amazonaws.com/api/v1"

Security Group

Port 30001 TCP/UDP is used by geth client, 9000 TCP/UDP is used by lighthouse for p2p communication, all of them are adviced to be open to everyone to get as more peers as possible, check lighthouse doc for more info.

You'll have to add rules to your node group's security group to allow these traffic, check this for how to do it manually.

Acknowledgement

This chart is developed based on vulcanlink's repo.