We have a Google managed Kubernetes cluster inside a subnetwork subA of VPC Network net and then we also have a subnetwork subB different region as part of that VPC. The Kubernetes cluster has VPC alias enabled so the service range and pod range are also practically subnetworks of the VPC (same local space

We want to make a kubernetes service (with an internal load balancer) hosted on subA and reachable by subB. We've managed to make the instances on subA and subB reachable to each other with firewall rules but those target only a network (not a subnetwork), service accounts, or tags, therefore cannot be used against load balancers.

Any clues?


As you say, subA and SubB subnetworks are part of the same VPC network, so there is no need to apply firewall rules in order to manage the traffic inside the VPC network between both subnets. The firewall only deals with the traffic that comes into (ingress) the VPC network from other networks and eventually the outgoing traffic (egress) that comes out of the VPC network.

It would be interesting to see a description (kubectl describe svc ) of your load balancer to check if it leads to the proper endpoints, this to make sure that a pod will respond to the requests done to the load balancer.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.