Skip to main content
Version: 3.6 (unsupported)
    Features and pricing

    Guidelines for creating an EKS cluster for ScalarDB Cluster

    This document explains the requirements and recommendations for creating an Amazon Elastic Kubernetes Service (EKS) cluster for ScalarDB Cluster deployment. For details on how to deploy ScalarDB Cluster on an EKS cluster, see Deploy ScalarDB Cluster on Amazon EKS.

    Before you begin​

    You must create an EKS cluster based on the following requirements, recommendations, and your project's requirements. For specific details about how to create an EKS cluster, see the official Amazon documentation at Creating an Amazon EKS cluster.

    Requirements​

    When deploying ScalarDB Cluster, you must:

    • Create the EKS cluster by using Kubernetes version 1.21 or higher.
    • Configure the EKS cluster based on the version of Kubernetes and your project's requirements.

    Recommendations (optional)​

    The following are some recommendations for deploying ScalarDB Cluster. These recommendations are not required, so you can choose whether or not to apply these recommendations based on your needs.

    Create at least three worker nodes and three pods​

    To ensure that the EKS cluster has high availability, you should use at least three worker nodes and deploy at least three pods spread across the worker nodes. You can see the sample configurations of podAntiAffinity for making three pods spread across the worker nodes.

    note

    If you place the worker nodes in different availability zones (AZs), you can withstand an AZ failure.

    Use 4vCPU / 8GB memory nodes for the worker node in the ScalarDB Cluster node group​

    From the perspective of commercial licenses, resources for one pod running ScalarDB Cluster are limited to 2vCPU / 4GB memory. In addition to the ScalarDB Cluster pod, Kubernetes could deploy some of the following components to each worker node:

    • ScalarDB Cluster pod (2vCPU / 4GB)
    • Envoy proxy (if you use indirect client mode or use a programming language other than Java)
    • Your application pods (if you choose to run your application's pods on the same worker node)
    • Monitoring components (if you deploy monitoring components such as kube-prometheus-stack)
    • Kubernetes components
    note

    You do not need to deploy an Envoy pod when using direct-kubernetes mode.

    With this in mind, you should use a worker node that has at least 4vCPU / 8GB memory resources and use at least three worker nodes for availability, as mentioned in Create at least three worker nodes and three pods.

    However, three nodes with at least 4vCPU / 8GB memory resources per node is the minimum for production environment. You should also consider the resources of the EKS cluster (for example, the number of worker nodes, vCPUs per node, memory per node, ScalarDB Cluster pods, and pods for your application), which depend on your system's workload. In addition, if you plan to scale the pods automatically by using some features like Horizontal Pod Autoscaling (HPA), you should consider the maximum number of pods on the worker node when deciding the worker node resources.

    Configure Cluster Autoscaler in EKS​

    If you want to scale ScalarDB Cluster pods automatically by using Horizontal Pod Autoscaler, you should configure Cluster Autoscaler in EKS too. For details, see the official Amazon documentation at Autoscaling.

    In addition, if you configure Cluster Autoscaler, you should create a subnet in an Amazon Virtual Private Cloud (VPC) for EKS with the prefix (e.g., /24) to ensure a sufficient number of IPs exist so that EKS can work without network issues after scaling.

    Create the EKS cluster on a private network​

    You should create the EKS cluster on a private network (private subnet in a VPC) since ScalarDB Cluster does not provide any services to users directly via internet access. We recommend accessing ScalarDB Cluster via a private network from your applications.

    Restrict connections by using some security features based on your requirements​

    You should restrict unused connections in ScalarDB Cluster. To restrict unused connections, you can use some security features in AWS, like security groups and network access control lists.

    The connections (ports) that ScalarDB Cluster uses by default are as follows:

    • ScalarDB Cluster
      • 60053/TCP (accepts gRPC or SQL requests from a client)
      • 8080/TCP (accepts GraphQL requests from a client)
      • 9080/TCP (accepts monitoring requests)
    • Scalar Envoy (used with ScalarDB Cluster indirect mode)
      • 60053/TCP (load balancing for ScalarDB Cluster)
      • 9001/TCP (accepts monitoring requests for Scalar Envoy itself)
    note
    • If you change the default listening port for ScalarDB Cluster in the configuration file (scalardb-cluster-node.properties), you must allow connections by using the port that you configured.
    • You must also allow the connections that EKS uses itself. For more details about Amazon EKS security group requirements, refer to Amazon EKS security group requirements and considerations.