Yeah almost all regions and zones for any compute instance have been exhausted since about 1pm PST on Friday. I finally got one up last night on us-east1, but my older cluster is basically SOL until it's fixed on us-west1. It went down for an upgrade and never came back up because of the same resource issue.