I am trying to formulate an alert to trip if CPU/Memory exceeds 95th percentile line for 5 min...anyone have a clue on the sql statement or if what I am looking for is even possible?
Wouldn't it make more sense to use a set % utilization instead? Otherwise you will get alerts that say things like CPU reached 21%.
What if it were for something like response time instead of CPU/Memory? For example, we wanted an alert if response time suddenly increased at a site. Setting up an alert trigger when it exceeds 95th percentile would have been a nice option. Instead, I created a dozen different "bucket" alerts and added sites that fit the appropriate bucket (alert when response time > 40ms, 50ms, 100ms, 200ms, etc).
actually no...many users have systems running at various CPU levels...take sql 2008 for example...it loves to run high, yet printer, dc, and other systems are low...then you have some odd ball boxes that may hover around 20-30
I want to configure the same alert. Can anyone assist here?
The aim is to get noticed about SPIKE/DIP on uplinks.
