I’m facing a bug using the backendconfig CRD on GCP.
I deployed a multicluster ingress and multi-cluster service and backendconfig CRD to configure the health check for the Loadbalancer, everything works fine except that GCP is not considering the specifications in the backend config file.
Above is the config file, but in the GCP console, the path is defined as / and the port is defined as 80, which makes the application fail a health check.
I did try to edit the health check from the console to the desired configuration, then the application works fine, but after some few minutes GCP returns to the wrong configuration.
Your question lack details such as the error message when you mentioned that it returns to the wrong configuration,
or, if you have followed any steps or guides, or any related documentation prior to the ingress creation.
You may try to reconfigure the health check by following this guide regarding health check configuration.
If the console is returning or giving an error message related to the backend config, ingress, etc, you may find the troubleshooting steps on this documentation.
Also if you observe my configuration aligns with the above documentation, including the one you sent. There is no specific error message but the health check configuration in the GCP console does not obey the CRD config which cause the health check to fail and the application returns the below error.
Make sure it is the container port, by the way. If your application inside the container serves on port 8080, that port needs to go into the backend config.
