Support for temporary http status codes

Use-Case: Someone manually puts an endpoint into some sort of maintenance mode so instead of return a 200 it now returns a 302.

Right now this triggers a status code metric to change for 0 to 1. It would be nice to be able to specify a grace period for this temporary status code.

Another option would be to output a different metric value when the temporary status code is being returned.

This would then be able to not trigger alarms unless the temporary status code remains in-place for greater than grace period

Start by tracing the Lambda's HTTP status handling and CloudWatch metric emission. Clarify whether the expected behavior is a configurable grace period or a different metric value for temporary status codes; done should define the configuration and ensure alarms do not trigger while the temporary response remains within the grace period.

Written by the indexing model from the issue text.