Those are two separate links to two separate incidents.

That was a Global Load Balancer outage. Not a GKE outage.

When that GCLB outage happened, GKE would still be able to get traffic if you used ngnix ingress and/or regional load balancers.