Deploy Domino on AWS Elastic Kubernetes Services (EKS)

This topic describes how to deploy Domino components on Amazon Elastic Kubernetes Service (EKS). EKS is hosted on Amazon Web Services (AWS).

Configure the cluster

Amazon Web Services Command Line Interface (AWS CLI) doesn’t support SOCKS5 proxies, so you must run the commands directly from a machine that has network access to the Kubernetes cluster.

Note	Before installing the following binaries, check for existing versions.

To connect to the bastion host, run the command generated by the Terraform output:
- If you deployed your infrastructure using the terraform-aws-eks module version: v3.0.1 or above you can use the included script tf.sh.
  ./tf.sh infra output ssh_bastion_command
- Otherwise:
  terraform output ssh_bastion_command

Verify that the following binaries are installed:

# kubectl
kubectl version --client=true --short=true

#aws-cli
aws --version

#docker daemon is installed and running.
docker --version
docker ps

Otherwise follow these steps to install those that are missing:

Update aws-cli to version 2.

curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" -o "awscliv2.zip"
unzip awscliv2.zip
sudo ./aws/install --bin-dir /usr/local/bin --install-dir /usr/local/aws-cli --update && rm -fr aws awscliv2.zip

Install Docker and exit the terminal. When you log in again, the modifications that you just made will become active:
```
sudo yum install -y docker
sudo systemctl start docker
sudo usermod -a -G docker ec2-user
exit
```

Install kubectl:

curl -LO https://dl.k8s.io/release/v1.24.10/bin/linux/amd64/kubectl
chmod +x kubectl
sudo mv kubectl /usr/local/bin/

Important

The kubectl version you install must be within one minor version difference of your EKS control plane. For example, if your EKS clusters are on Kubernetes version 1.21, install a version 1.20, 1.21, or 1.22 kubectl client.

Use environment variables to set the values of IDs, names, and labels. This simplifies the commands you’ll run while installing Domino components:

unset HISTFILE
export DOMINO_VER=<The Domino version to deploy>
export FLEET_COMMAND_TAG=<Tag that corresponds to the version of Domino deployed>
export DEPLOY_NAME=<Name of deployment>
export AWS_REGION=<The region to deploy the resources>
export AWS_ACCESS_KEY_ID=<Your AWS access key ID>
export AWS_SECRET_ACCESS_KEY=<Your AWS secret key>
export QUAY_USERNAME=<`quay.io` username provided by Domino>
export QUAY_PASSWORD=<`quay.io` password provided by Domino>

Tip	You can get the deployment tag from the fleetcommand-agent release notes.

To retrieve the credentials for your Kubernetes cluster, check your local kubeconfig with:
```
export KUBECONFIG=$(pwd)/kubeconfig
```

Update the kubeconfig:

aws eks update-kubeconfig --kubeconfig $KUBECONFIG --region $AWS_REGION --name $DEPLOY_NAME --alias $DEPLOY_NAME

If you aren’t already logged into quay.io, run the following:

docker login -u $QUAY_USERNAME -p $QUAY_PASSWORD quay.io

Generate a configuration file named domino.yml in your working directory:
```
docker run --rm -it
  -v $(pwd):/install
  quay.io/domino/fleetcommand-agent:$FLEET_COMMAND_TAG
  init
  --file /install/domino.yml --version $DOMINO_VER
  --preset eks
```
Caution

This overwrites existing domino.yml files.

Changing the defaults in domino.yml can affect the deployment. If you must adjust its parameters, contact a Domino representative.
Gather Terraform values to update the domino.yml file:
- If you deployed your infrastructure using the terraform-aws-eks module version: v3.0.1 or above you can use the included script tf.sh.
  ./tf.sh infra output domino_config_values
- Otherwise:
  terraform output domino_config_values

Open the domino.yml file and edit the attributes as follows:

name: The name of the deployment. This can’t be changed post-deployment.
hostname: The hostname for the Domino install (for example, domino.example.com).
autoscaler.auto_discovery.cluster_name: Name of the k8s cluster.

autoscaler.aws.region: AWS deployment region.

Note	Configure only one of the following: either `external_docker_registry` or `internal_docker_registry`. The `external_docker_registry` should only be set during new installations. If you are upgrading and have previously configured `internal_docker_registry`, you must continue to use it.

external_docker_registry: Specifies the ECR Repository URL.
internal_docker_registry.s3_override.region: AWS region for the S3 registry bucket.
internal_docker_registry.s3_override.bucket: S3 bucket name for internal docker registry.
internal_docker_registry.s3_override.sse_kms_key_id: KMS key for S3 internal docker registry bucket.
storage_classes.block.parameters.kmsKeyId: KMS key for block storage.
storage_classes.shared.efs.region: AWS region for the EFS system.
storage_classes.shared.efs.filesystem_id: EFS filesystem id.
storage_classes.shared.efs.access_point_id: EFS access point id.
blob_storage.projects.s3.region: S3 bucket region for projects.
blob_storage.projects.s3.bucket: S3 bucket name for projects.
blob_storage.projects.s3.sse_kms_key_id: KMS key for S3 projects bucket.
blob_storage.logs.s3.region: S3 bucket region for logs.
blob_storage.logs.s3.bucket: S3 bucket name for logs.
blob_storage.logs.s3.sse_kms_key_id: KMS key for S3 logs bucket.
blob_storage.backups.s3.region: S3 bucket region for backups.
blob_storage.backups.s3.bucket: S3 bucket name for backups.
blob_storage.backups.s3.sse_kms_key_id: KMS key for S3 backups bucket.

Configure the LoadBalancer by adding the following code to the end of the file. Replace <SSL certificate arn> and <Monitoring bucket name> with the values for your installation.

release_overrides:
  nginx-ingress:
    chart_values:
      controller:
        kind: Deployment
        hostNetwork: false
        service:
          enabled: true
          externalTrafficPolicy: "Local"
          type: LoadBalancer
          annotations:
            service.beta.kubernetes.io/aws-load-balancer-ssl-negotiation-policy: 'ELBSecurityPolicy-TLS-1-2-2017-01'
            service.beta.kubernetes.io/aws-load-balancer-ssl-cert: '<SSL certificate arn>'
            service.beta.kubernetes.io/aws-load-balancer-type: 'nlb'
            service.beta.kubernetes.io/aws-load-balancer-internal: 'false'
            service.beta.kubernetes.io/aws-load-balancer-backend-protocol: 'ssl'
            service.beta.kubernetes.io/aws-load-balancer-ssl-ports: '443'
            service.beta.kubernetes.io/aws-load-balancer-connection-idle-timeout: '3600'
            service.beta.kubernetes.io/aws-load-balancer-cross-zone-load-balancing-enabled: 'true'
            service.beta.kubernetes.io/aws-load-balancer-access-log-enabled: 'true'
            service.beta.kubernetes.io/aws-load-balancer-access-log-emit-interval: '5'
            service.beta.kubernetes.io/aws-load-balancer-access-log-s3-bucket-name: '<Monitoring bucket name>'
            service.beta.kubernetes.io/aws-load-balancer-access-log-s3-bucket-prefix: 'ELBAccessLogs'
          loadBalancerSourceRanges:
            - 0.0.0.0/0 # should always be 0.0.0.0/0, networkPolicy does the enforcement
        config:
          use-forwarded-headers: 'false'
          ssl-ciphers: 'ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-SHA256:ECDHE-ECDSA-AES128-SHA256:ECDHE-RSA-AES128-SHA256:ECDHE-ECDSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-ECDSA-AES256-SHA384:ECDHE-RSA-AES256-SHA384:AES128-GCM-SHA256:AES128-SHA256:AES256-GCM-SHA384:AES256-SHA256:!aNULL:!eNULL:!EXPORT:!DES:!RC4:!MD5:!PSK:!aECDH:!EDH-DSS-DES-CBC3-SHA:!EDH-RSA-DES-CBC3-SHA:!KRB5-DES-CBC3-SHA'
          ssl-protocols: 'TLSv1.2 TLSv1.3'
      networkPolicy:
        ipCidrs:
          - 10.0.0.0/16   # node network
          - 100.64.0.0/16 # pod network
          - 0.0.0.0/0     # access list starts here

Note

The CIDR 0.0.0.0/0 for loadBalancerSourceRanges can be updated to restrict access to certain CIDR blocks.
We have switched to Network Load Balancers. Existing installs should either continue to use the existing Classic Load Balancer configuration or otherwise be prepared to adjust DNS for the replacement Network Load Balancer on upgrade, as Kubernetes will destroy the currently provisioned Classic Load Balancer when changing the configuration.

For reference, this is the old Classic Load Balancer configuration:

release_overrides:
  nginx-ingress:
    chart_values:
      controller:
        kind: Deployment
        hostNetwork: false
        service:
          enabled: true
          type: LoadBalancer
          annotations:
            service.beta.kubernetes.io/aws-load-balancer-ssl-negotiation-policy: ELBSecurityPolicy-TLS-1-2-2017-01
            service.beta.kubernetes.io/aws-load-balancer-ssl-cert: <SSL certificate arn>
            service.beta.kubernetes.io/aws-load-balancer-internal: false
            service.beta.kubernetes.io/aws-load-balancer-backend-protocol: tcp
            service.beta.kubernetes.io/aws-load-balancer-ssl-ports: '443'
            service.beta.kubernetes.io/aws-load-balancer-connection-idle-timeout: '3600'
            service.beta.kubernetes.io/aws-load-balancer-proxy-protocol: '*'
            service.beta.kubernetes.io/aws-load-balancer-access-log-enabled: 'true'
            service.beta.kubernetes.io/aws-load-balancer-access-log-emit-interval: '5'
            service.beta.kubernetes.io/aws-load-balancer-access-log-s3-bucket-name: <Monitoring bucket name>
            service.beta.kubernetes.io/aws-load-balancer-access-log-s3-bucket-prefix: ELBAccessLogs
          loadBalancerSourceRanges:
            - 0.0.0.0/0
          targetPorts:
            http: http
            https: http
        config:
          use-proxy-protocol: 'true'

Install Domino with `fleetcommand-agent`

fleetcommand-agent installs and configures Domino components. It uses the installation template to gather the required parameters for the environment and sets them when installing Domino components.

To install Domino on the infrastructure you prepared, run the following:

curl -o fleetcommand-agent-install.sh https://docs.dominodatalab.com/attachments/fleetcommand-agent-install.sh
bash fleetcommand-agent-install.sh $DOMINO_VER

See fleetcommand-agent-install.sh Downloads for more information.

Tip	If you encounter errors, investigate and resolve the root cause before you run `fleetcommand-agent-sh` again. Failures are often related to resource quotas and limits. Contact a Domino representative for assistance.

Set up DNS

Run the following to get the external IP to access your instance’s Domino management plane:

kubectl -n domino-platform get svc nginx-ingress-controller

You can use this to update your DNS records accordingly.

Create a network policy if you use your own ingress controller

If you use your own NGINX ingress controller by specifying ingress_controller.install = false, then you need to create a network policy in the Domino platform and compute namespace.

Here is an example of a network policy that allows ingress from the nginx namespace:

kubectl -n <domino-namespace> apply -f - <<EOF
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: external-nginx
spec:
  ingress:
  - from:
    - namespaceSelector:
        matchLabels:
          kubernetes.io/metadata.name: nginx
  podSelector: {}
  policyTypes:
  - Ingress
EOF

Validate your installation

Go to https://<YOUR-DOMAIN>/auth/.
Login with the username keycloak and the password from the keycloak-http secret in the domino-platform namespace.

Use the following command to get the password:

echo -e "
your password is: $(kubectl get secret keycloak-http  -n domino-platform --template={{.data.password}} | base64 -d)
"

Go to Users in the navigation pane and click Add User.
Enter the username, first name, last name, and email address, and then click Save.
Go to the Credentials tab and add a password.
Optional: Disable Temporary.
Click Set Password.
Go to Role Mappings.
From Client Roles, select domino-play.
Select the User role and add it to your user.
Go to the main page for your Domino deployment (for example, https://<YOUR-DOMAIN>) and sign in with your new Domino user.
Go to Environments > Domino Standard Environment Py3.8 R4.1 > Revisions and make sure the revision is active. If not, use Build Logs to try to solve the problem.
Go to Projects > Quick-start > Workspaces and launch a new workspace using Jupyter (this can take a while).
When the new workspace is created open main.ipynb and confirm that you can execute the script without errors.

Enable user registration

Use Keycloak to enable user registration, so users can access your fresh Domino install. Keycloak is a user authentication service that runs on a pod in your cluster.