One of my GKE clusters is running out of IPs for Services. It was originally created with the default service address range (10.0.0.0/20), and we are about to hit the 4k services mark.
Is there a way to expand the service address range of an existing GKE cluster? The cluster's subnet has enough space to allocate more IPs, but I haven't found an API or command to change it.
https://cloud.google.com/kubernetes-engine/docs/how-to/alias-ips#cluster_sizing_secondary_range_svcs mentions that editing the secondary range is not possible while it's in use. Is there a way to replace it, and let GKE roll the IPs of the services?

