Have you checked the health monitor to make sure it triggers instantly? Generally, you want some leniency to make sure you don't generate false alerts. 

Well currently I'm just going in and marking a pool member as "disabled/down", so I'm not even using the health check to mark it as down.   That event shows up under Operations/Events/All Events instantly,  however its around 2 minutes before I receive an email telling me that the backend server is down.  Also coincidentally ,  it also takes about 2 minutes before I receive a syslog event in Splunk that the backend server is down.  I feel like this alert should happen instantly.

Thank you,

Tony

Gotcha. Only other config item I can think of is that the alert itself might be set for "Rolling window" instead of instance. It might be worth a check, my personal next stap would be to contact support and ask them