I have a health check with a 1-second check interval (http://s.drollette.com/0B2A3Z1w2X1G). It was created by the GLBC Ingress controller in kubernetes. But looking at the logs it appears that it is generating 3 requests per second (http://s.drollette.com/2U432C2f1d2f). Is this expected behavior from a Google Compute Health Check? Nothing else is configured to be hitting that route.
Yes, this is the expected behavior. More than one health checker is used to check the service, and each health checker independently obeys the interval specified in your configuration.