Monitor and Troubleshoot
When your platform runs an observability plane, OpenChoreo collects the logs, Kubernetes events, metrics and traces of your components, raises the alerts you define, and can open incidents and AI root cause analysis reports for them. You read all of it by OpenChoreo concepts (project, component, environment and release line), without knowing where the data is stored.
What You Can Use
| What | Backstage portal | CLI | See |
|---|---|---|---|
| Logs | Logs tab on the component and the project | occ component logs | Logs and Events |
| Kubernetes events | Events tab on the component | - | Logs and Events |
| Resource and HTTP metrics | Metrics tab on the component and the project | - | Metrics |
| Traces | Traces tab on the project | - | Traces |
| Alerts | Alerts tab on the component | - | Alerts and Incidents |
| Incidents and RCA reports | Incidents and RCA Reports tabs on the project | - | Alerts and Incidents |
| Network flows (wirelogs) | Wirelogs tab on the component | - | Troubleshooting |
| Deployment and resource status | Deploy tab and View K8s Artifacts on the component | occ component get | Troubleshooting |
The Observability Plane MCP server gives AI assistants the same logs, events, metrics, traces, alerts and incidents.
Before You Start
- An observability plane. Your platform engineer installs it and links it to the data plane of each environment. Without one, the portal tabs say that observability is not enabled for the component in that environment. Which signals are available also depends on the modules your platform runs: for example, events need an events collector, traces need a tracing module, and HTTP metrics need the Cilium network module.
- Permissions. The default
developerrole can read logs, events, metrics, traces, alerts and RCA reports. Incidents and wirelogs need permissions it doesn't include. Ask your platform engineer if a tab shows Permission Denied.
Common Filters
Most tabs share the same filters:
- Environment: each environment of the project's deployment pipeline.
- Time range: the last 10 minutes up to the last 30 days, or a custom range.
- Release line: for a component with release lines, all lines, the main line or one named line. It starts on the main line.
What's Next
- Logs and Events: search your component's logs and Kubernetes events
- Alerts and Incidents: define alert rules on your component
- Troubleshooting: inspect resources and fix common issues