The possible cause is that the Gateway-related resource configurations are missing or incorrect. Do as follows to locate the fault:
- Log in to the Elastic Load Balance console, check whether the external port and ECSs are successfully listened by the load balancer.
- Log in to the cluster and run the kubectl get gateway -n istio-system command to check whether the IP address, domain name, and port number are configured for the Gateway. Run the kubectl get svc -n istio-system command to check whether the ingress Gateway has the corresponding IP address and port and is not in the pending status.
- Check whether the internal access protocol of the service added to the service mesh is consistent with the external access protocol configured for the service's Gateway.
- If the ERR_UNSAFE_PORT error is displayed when you use a browser to access the service, that is because the port is identified as a dangerous port by the browser. In this case, you need to use another external port.