GKE import stuck in Pending

I followed all the instructions but the cluster shows as Pending and the import never finishes. If I use the GKE console, I can see the cattle pods are deployed to the cluster but they are in a state of CrashLoopBackOff. When I go to the logs for those pods in the GKE console, they are empty. There doesn’t appear to be anywhere in the Rancher UI to see logs or anything else as to what is wrong. Any suggestions?

I’ve made some progress. I found the logs on the pod’s VM in /var/log/containers. The error is: “Failed to pull the cacert from the rancher server.” This gives me a place to start but the 2.0 documentation is fairly thin so if anyone knows how to fix this easily I’d appreciate a nudge in the right direction.

It turned out this was a GCE firewall rule issue. An Ingress rule was needed on the Rancher host to allow traffic from the GKE nodes. Once the rule was added, the cluster became active in Rancher.