Deploy OpenSearch

Learn how to deploy OpenSearch as a search and analytics engine on your Kubernetes cluster using lowcloud.

Last updated: March 16, 2026

OpenSearch is an open-source, distributed search and analytics engine (Apache 2.0). It is ideal for log analytics, real-time monitoring, full-text search, and security analytics. With the official Helm Chart, you can deploy OpenSearch including OpenSearch Dashboards in your Kubernetes cluster.

Chart Registry: https://opensearch-project.github.io/helm-charts/

Prerequisites

A lowcloud account with an active cluster
Access to the lowcloud Dashboard

Quick Start

Minimal configuration to get OpenSearch up and running:

singleNode: true

resources:
  requests:
    cpu: 250m
    memory: 512Mi

persistence:
  enabled: true
  size: 10Gi

For development and testing environments, singleNode: true is sufficient. For production, a multi-node setup with dedicated master and data nodes is recommended.

Configuration

Overview of the most important configuration options:

Parameter	Description	Default
`singleNode`	Enable single-node mode	`false`
`replicas`	Number of OpenSearch nodes	`3`
`resources.requests.cpu`	CPU request per node	`1000m`
`resources.requests.memory`	Memory request per node	`512Mi`
`persistence.enabled`	Enable persistent storage	`true`
`persistence.size`	Size of the persistent volume	`8Gi`
`opensearchJavaOpts`	JVM options (e.g. heap size)	`-Xmx512M -Xms512M`
`plugins.security.disabled`	Disable security plugin	`false`

Step-by-Step Guide

1. Add a Service

Navigate to Add Service in the lowcloud Dashboard and select Helm Release. Add the official OpenSearch registry as the chart repository:

https://opensearch-project.github.io/helm-charts/

Select the opensearch chart and then the desired version. The "latest" option is a good choice in most cases.

Select Helm Chart version

2. Configure Values

Give the Helm Release a descriptive name in lowcloud (e.g. opensearch-logs).

Then adjust the values.yaml to your requirements:

singleNode: true

resources:
  requests:
    cpu: 250m
    memory: 512Mi
  limits:
    cpu: 1000m
    memory: 1Gi

persistence:
  enabled: true
  size: 10Gi

opensearchJavaOpts: '-Xmx512M -Xms512M'

Adjust opensearchJavaOpts to match the available memory resources. The heap size should be approximately 50% of the memory limit.

Configure Helm Release

3. Start Deployment

Open the created service and click Deploy. Wait until the label in the Dashboard changes to "Deployed". Your OpenSearch instance is then ready to use.

You can now access OpenSearch via the REST API to index, search, and analyze data.