The RHEL CoreOS team has sanity tests for cri-o that started failing when we moved to cri-o 1.18

Our previous configuration (pod template/container template) no longer works with 1.18. We are getting failures like:

            cluster.go:141: time="2020-04-20T18:01:00Z" level=fatal msg="run pod sandbox failed: rpc error: code = Unknown desc = setting pod sandbox name and id: cannot generate pod name without uid in metadata"

The error seems to match 2d5ceda

I suspect we'll be told to update our templates, but filing an issue to track the issue going from 1.17 to 1.18 for us.