Multi-node deployment

This page provides instructions for installing, updating, and uninstalling the Aindo Synthetic Data Platform in a multi-node Kubernetes cluster. Based on the architecture outlined in the dedicated section, a multi-node deployment can be represented as follows:

In the sections below, we assume that the Kubernetes cluster is already operational and that the kubectl command-line tool is properly configured.

Nodes setup

Cluster nodes should be organized into two distinct pools:

1. a pool for serving the web application
2. a pool for running synthesis workers

To simplify the deployment process, it is essential to ensure the following:

• nodes should have at least one Kubernetes label to indicate the pool they belong to (in managed clusters, this label is often assigned by default and typically includes the node pool name). For example, use workload=generic for web application nodes and workload=worker for worker nodes.
• nodes dedicated to synthesis workers should have a NoSchedule taint to prevent unintended workloads. An example taint configuration might be: key: workload, value: worker, effect: NoSchedule.

Installation

Step 1 - Install Helm

The first step is to install Helm, a package manager for Kubernetes. For installation instructions, refer to the official documentation.

Step 2 - Log in to the registry

The Aindo Synthetic Data Platform Helm chart is hosted in a private OCI registry on Docker Hub. Log in using:

helm registry login -u aindo registry-1.docker.io

To confirm access to the Helm chart, run:

helm show readme oci://registry-1.docker.io/aindo/aindo-data-platform --version 1.3.1

and you should see the Helm chart README.

Step 3 - Deploy the Aindo Platform

Prepare the Helm values file with the appropriate configuration for your deployment. Assuming the file is named values.yaml and the target namespace is aindo, use the following command to deploy the platform:

helm install aindo-data-platform -f values.yaml -n aindo --create-namespace helm-chart-path/ oci://registry-1.docker.io/aindo/aindo-data-platform --version 1.3.1

To verify that the deployment was successful, run:

helm status aindo-data-platform -n aindo

Expected output:

NAME                NAMESPACE   REVISION   UPDATED                        STATUS    CHART                      APP VERSION
aindo-data-platform aindo       1          2025-02-05 16:28:16 +0100 CET  deployed  aindo-data-platform-1.3.1  X.Y.Z

Updating

To upgrade the platform, run:

helm upgrade aindo-data-platform -f values.yaml -n aindo oci://registry-1.docker.io/aindo/aindo-data-platform --version 1.3.1

Uninstallation

To uninstall the platform, use the following command:

helm uninstall aindo-data-platform -n aindo