Feature Suggestion - Node Health

We all know we can make an alert on just about anything, especially when you bring custom pollers into the mix.  But I think we can take that a step further.

I find myself using alerts and different views to access node health.  I was thinking would it not be so much easier and more accurate if Node Health was a feature on its own.

The way I see this working is you create a node health profile.  In this profile you use any of the usual node/interface details or SQL statements that you would normally use to filter a view with and use that to apply to the nodes you want to monitor in this health node profile.

Then once you have the nodes selected you want this to apply to you add the specific health high water marks that an item (CPU, RAM etc) would need to be under to consider this node healthy.

Then alerts could be created to go off when a node is not healthy and that level would be configurable IE node is unhealthy if 2 of the 10 items are over the limit.

Reports too could be made, IE show me all the nodes that are currently not health or have gone not health X times.

Obviously this could be applies to servers, routers, switches and the like all with different profiles specific to that node type.

The goal of this for me would be to allow me to be as proactive as possible with the least amount of screen time.