Overview
- Install the argo-events namespace and Argo Events manifests.
- (Optional) Install Argo Workflows if you plan to trigger workflows from events.
- Verify CRDs, RBAC, config maps, and the controller deployment.
- Inspect pods and logs to ensure the controller becomes Ready.
- Use the Argo Workflows UI to view EventSources and Sensors when integrated.
Prerequisites
- A Kubernetes cluster with kubectl configured to target the cluster.
- (Optional) If you plan to use the Argo Workflows UI, install Argo Workflows into the cluster.
Installation
Create the argo-events namespace and apply the official Argo Events install manifest:Verify installation
List all resources in theargo-events namespace:
ContainerCreating until the container is ready. Example output while the controller is still starting:
UI: Viewing EventSources and Sensors
Argo Events does not include a standalone web UI. If you install Argo Workflows and integrate it with Argo Events, the Argo Workflows UI can display EventSources and Sensors. Open the Argo Workflows UI and select the appropriate namespace (for example,argo-events) to view event-related resources.



If the controller pod remains in ContainerCreating or enters CrashLoopBackOff, inspect the pod events and logs with:
kubectl -n argo-events describe pod <pod-name>kubectl -n argo-events logs <pod-name>These commands help identify scheduling, image pull, or runtime errors.
Common resources created by the install
| Resource Type | Purpose |
|---|---|
| CustomResourceDefinitions (CRDs) | Define EventBus, EventSource, and Sensor custom resources |
| ServiceAccount, ClusterRole, ClusterRoleBinding | RBAC for the controller to watch and manage resources |
| ConfigMap | Controller configuration (e.g., metrics, event bus settings) |
| Deployment / ReplicaSet / Pod | Controller manager that reconciles EventSources and Sensors |
Troubleshooting checklist
- Confirm the controller pod is scheduled and not pending:
kubectl -n argo-events get pods
- If
ContainerCreating, check node space and image pull errors:kubectl -n argo-events describe pod <pod-name>
- If the pod crashes repeatedly, view logs for stack traces:
kubectl -n argo-events logs <pod-name>
- Verify CRDs were created successfully:
kubectl get crd | grep argo
Next steps
- Wait for the
controller-managerpod to reach READY state before creating EventSources and Sensors. - Create EventSource and Sensor manifests to connect your external events to triggers (e.g., Webhook, cron, Kafka) and test end-to-end behavior.
- If using Argo Workflows, create a Workflow template and configure a Sensor trigger to submit workflows on events.
Links and references
- Argo Events (project): https://argoproj.github.io/argo-events/
- Argo Workflows (project): https://argoproj.github.io/argo-workflows/
- Kubernetes documentation: https://kubernetes.io/docs/