OpenChoreo deployment succeeds but service is not accessible via generated endpoint

I hv run into similar issues before… When the deployment succeeds and the pods are healthy, a connection refused error often points to the application listening on a different port than the one configured in OpenChoreo. That’s usually the first thing I’d verify by checking the application startup logs and the component’s port configuration.

I would also check the Kubernetes Service and Endpoints to make sure traffic is being forwarded to the container port. If this is possible I will try port forwarding directly to the pod.

If the application is reachable through port forwarding but not through the generated endpoint the problem is likely in the Kubernetes Service, the ingress or the gateway layer than the Kubernetes application itself.

Another useful place to look is the ingress logs or the gateway logs along with any OpenChoreo deployment events.

The Kubernetes ingress logs and the gateway logs often provide context than the Kubernetes application logs alone. They can help identify routing problems or networking misconfigurations, in the Kubernetes Service.

Thanks! Looks like I have a few more places to investigate before blaming OpenChoreo

I also ran into similar issues during deployments. It usually happens if the port is wrong or the bind IP is not 0.0.0.0 (some frameworks bind to 127.0.0.1 unless you specified). Then there is no one accepting the traffic coming from the gateway. Are you getting a 503 status when accessing the endpoint?

Another thing I noticed is that you said that your application logs show "connection refused" errors. Does it connect to another service or DB? If that is the case, this can be due to your app failing to listen because it is not connecting to the other service, depending on your implementation.