[New Scheduler] Healthcheck actions are created by invoker which scales linearly with fleet size

[bdoyle0182]

Is there any value in requiring each invoker create a healthcheck action that can't be achieved by just having a single healthcheck action referenced by all invokers? Previously the healthcheck actions were managed by the controllers which doesn't scale with the amount of traffic your cluster manages, but now that is no longer true on the new scheduler where there is a duplicate function for each healthcheck. If it's not necessary, it should be a super quick change to just have a single healthcheck function key that is referenced by all invokers.