Deploy AutoObserve

Choose how to deploy AutoObserve — Kubernetes, Helm, Docker, or manual install.

When should I use this?

You are installing AutoObserve for the first time and need to pick a deployment model.
You want to compare Kubernetes, Helm, and Docker options before committing.
You are onboarding as a design partner with a guided cluster setup.

Requirements

Kubernetes cluster or Docker environment for the chosen pathRequired
OpenTelemetry pipelines exporting metrics, logs, and tracesRequired
Access to deployment and change signals for correlationRequired

Steps

1. Choose a deployment option

Kubernetes

Recommended

Deploy AutoObserve into your cluster with Helm. Best for production and design-partner evaluations.

Difficulty: Intermediate
Time: ~15 min
Production readiness: Production ready

Helm

Production

Install with Helm when you already operate chart-based releases in Kubernetes.

Difficulty: Intermediate
Time: ~10 min
Production readiness: Production ready

Docker

Run AutoObserve locally or in a development cluster to explore the product.

Difficulty: Beginner
Time: ~5 min
Production readiness: Development only

2. Complete install and confirm access

Until public install artifacts ship, Design Partners get guided setup against real workloads.

Did it work?

Troubleshooting

What's happening?

Problem

Install fails or pods remain unready.

Symptoms

Helm install errors
Pods CrashLoopBackOff

Likely causes

Insufficient cluster resources
Missing secrets or chart values

Resolution

Review design-partner install notes or cluster events. Confirm secrets and resource limits.

Verify fix

All AutoObserve pods reach Running state.