Run MATLAB Parallel Server on Kubernetes in Amazon Web Services

Use a customizable reference architecture to run MATLAB^® Parallel Server™ on Kubernetes^® in Amazon^® Web Services (AWS^®) using Amazon Elastic Kubernetes Service (EKS).

After deploying this solution, you can run large-scale parallel computations without having to manage your own Kubernetes control plane. EKS provides elastic, on‑demand scaling for worker nodes, allowing MATLAB workloads to grow and shrink automatically based on demand.

Requirements

To use this reference architecture, you need:

• A MATLAB Parallel Server license. You can use either:

  • A MATLAB Parallel Server license configured to use online licensing for MATLAB. For information on how to configure your license for cloud use, see Configure MATLAB Parallel Server Licensing for Cloud Platforms (MATLAB Parallel Server).

  • A network license manager for MATLAB hosting sufficient MATLAB Parallel Server licenses for your cluster. MathWorks^® provides a reference architecture to deploy a suitable Network License Manager for MATLAB on Amazon Web Services (GitHub), or you can use an existing license manager.

• MATLAB and Parallel Computing Toolbox™ on your client machine.

• An AWS account with required permissions.

• AWS Command Line Interface (AWS CLI) installed on your client machine.

• An existing VPC with two subnets.

• AWS credentials configured on your client machine.

• Helm^® package manager version 3.8.0 or later installed on your client machine.

• kubectl command-line tool installed on your client machine and configured to access your Kubernetes cluster.

• Terraform or OpenTofu.

Run MATLAB Parallel Server on Kubernetes on Amazon EKS from GitHub

To deploy MATLAB Parallel Server on Kubernetes in AWS, see the instructions in this GitHub^® repository.

• MATLAB Parallel Server on Kubernetes in AWS (GitHub)

This reference architecture contains two main components:

1. Terraform or OpenTofu module: This module sets up all the required infrastructure in AWS, including EC2 instances, security groups, IAM roles and policies, networking components, autoscaling groups, and the EKS cluster itself.

2. Helm chart: This chart deploys MATLAB Parallel Server on the EKS cluster, including the job manager, workers, and all necessary Kubernetes resources. The Helm chart bundles two primary sub-charts, that you can configure using the Helm values file.

   • MATLAB Parallel Server in Kubernetes: Deploys the core MATLAB Parallel Server components. For more information, see Run MATLAB Parallel Server on Kubernetes (MATLAB Parallel Server).

   • Autoscaling: Configures autoscaling for worker nodes. For more information, see Cluster Autoscaler on AWS in the Kubernetes Autoscaler GitHub repository.