Very simple scenario StackDriver logging alert scenario. Alert me if there are more than 100 404s log entries in the nginx log in a 1 minute period. I've setup the log filter and this works correctly, filtering 404 lines.
And I can almost get the graph to show this in dashboard, and I guess the minor difference here is in the sampling and minute boundaries, that's ok, good enough.
But defining a policy I can't figure out how to specify this, as the policy graph is back to 0.x/sec, one can't specify sampling period or aggregation here.