Guide to the OpenTelemetry Kubernetes Operator that automates deploying and managing the OpenTelemetry Collector and automatic instrumentation in clusters, including installation and lifecycle management.
The OpenTelemetry Kubernetes Operator automates the deployment and lifecycle of the OpenTelemetry Collector and the auto-instrumentation layer inside a Kubernetes cluster. This guide explains how the operator works, what it manages, and how to get started with a minimal deployment.At a high level, the operator runs inside your cluster and manages two primary concerns:
The OpenTelemetry Collector (deployed as a Deployment, DaemonSet, or sidecar).
Automatic instrumentation injection that enables workloads to emit telemetry without changing application code.
When auto-instrumentation is enabled across namespaces, instrumented workloads start sending telemetry to the operator-managed collector. The collector then receives and routes that telemetry to your observability backend.
Helm (recommended for most users): simple, configurable, and integrates well with CI/CD.
Apply YAML manifests directly: useful for air-gapped or tightly controlled clusters.
Before installing, ensure a certificate manager is available in the cluster. The operator requires webhook TLS certificates to perform automatic injection and manage CRD webhooks.
cert-manager (or an equivalent certificate manager) must be present in the cluster before installing the operator because the operator creates webhook certificates automatically. See the cert-manager documentation for installation steps.
If you prefer not to use Helm, the operator repository provides YAML manifests that you can apply with kubectl apply -f. See the operator repo for the canonical manifests and instructions.
The operator can create and manage an OpenTelemetry Collector for you using the OpenTelemetryCollector custom resource. Below is a minimal example that deploys a collector in deployment mode with an OTLP receiver and a logging exporter.Save and apply this YAML after the operator is installed:
This CR instructs the operator to create a Kubernetes Deployment (or the selected mode) running the configured collector. You can extend the config block to add additional receivers, processors, exporters, and pipelines.
The OpenTelemetry Operator simplifies observability in Kubernetes by automating many operational tasks. Key responsibilities include:
Ports & Networking: create Kubernetes Services to expose the collector and manage configured ports.
Monitoring & Endpoints: register and manage collector endpoints so instrumented workloads can send telemetry across namespaces.
Upgrades & Lifecycle: handle upgrades, rollouts, restarts, and lifecycle events for operator-managed collectors and instrumentation artifacts.
Auto-instrumentation: configure and inject SDKs or sidecar instrumentations into workloads via webhooks.
Capability
What it does
Example
Ports & Networking
Creates Services and exposes configured ports for the collector
Service with OTLP port
Monitoring & Endpoints
Manages endpoints and cross-namespace telemetry flow
Configure OTLP receivers and routing
Upgrades & Lifecycle
Automates rolling updates and configuration rollouts
Operator-controlled Deployment updates
Auto-instrumentation
Performs injection for supported runtimes via webhooks
JVM/Node auto-instrumentation injection
Because the operator supports multiple deployment patterns (Deployment, DaemonSet, sidecar), automatic service creation, and webhook certificate management, it removes much of the manual effort required to run telemetry agents or instrument applications in Kubernetes.
The OpenTelemetry Operator simplifies deployment and lifecycle management of the OpenTelemetry Collector and auto-instrumentation for Kubernetes workloads.
Install via Helm (recommended) or by applying manifests directly. Ensure cert-manager (or equivalent) is present to provision webhook TLS certificates.
The operator automates service/port creation, endpoint management, rollouts, and automatic instrumentation injection.
