Improving the health check monitors by checking server performance

Hi guys,

Hoping you are all safe during these uncertain times!

I would like to ask for your expert opinion on a situation I experienced recently with one of my customers, related to slow response time from the backend servers on a HTTP monitor, which further causes slow response from the VIP back to the clients.

When verifying with application owners they observed that although the app is responding to healtcheck the CPU/memory skyrocket and that causes slowness in displaying the web page results.

Sometimes it even takes up to 1 minute for the whole page to display.

I was thinking about building a SNMP monitor to check the CPU/performance on the backend server and take a load balancing decision based on these values.

Does anyone have any ideas on how we can improve the application delivery time by best monitoring the backend servers based on CPU/memory or other apache triggers?

Thanks!