Common Issues
Stateless-lb-frontend containers are not ready
- Check the Gateways configuration.
- Misconfiguration of the Attractor
- Misconfiguration of the Gateways
All resources are not deployed with a Trench/Conduit/Attractor
- Does the custom resource name match the limitations (Trench / Conduit / Attractor).
- Check the logs in the Operator for errors.
Containers are not ready
- Check the health probes in the logs (Operator / TAPA / Stateless-lb / Frontend / NSP / IPAM / Proxy)
Traffic not received by any targets
- Is the stream status open in the targets.
- Find what path the traffic is using with the Data Plane documentation.
- Misconfiguration of the Flows / VIPs
- Faulty connectivity
Traffic received by the wrong targets
- Is the stream status open in the targets and closed in the wrong targets.
- Find what path the traffic is using with the Data Plane documentation.
- Misconfiguration of the Flows / VIPs
- Faulty connectivity
Traffic received by only a few targets
- Is the stream status open in all targets.
- Find what path the traffic is using with the Data Plane documentation.
- Misconfiguration of the Flows / VIPs
- Faulty connectivity
Some pods are losing connectivity with primary kubernetes network
- Check if the internal Meridio subnet is conflicting with the Kubernetes ones (Pod, Service, DNS).
metadata.labels.trench: Forbidden: update on label trench is forbidden
- A custom object cannot be switched between trenches. Check the limitations (Conduit / Stream / Flow / VIP / Attractor / Gateway)
Failed to call webhook: connect: connection refused
- The operator is not fully ready.
Stream status is unavailable
in the TAPA
- The NSP service is unreachable.
- The NSP pod is not running (Upgrade, Not ready...).
- The namespace is misconfigured.
- Too many targets are registered in the stream.
- The nsp service name/port is misconfigured.
Stream status is undefined
in the TAPA
- The stream/conduit/trench is not existing.
Something else
Please, open an issue on Github.