Here is an interesting one to think about.
I need to look into a way of adding advanced logic to "Thin AP Down" alerts to create a smarter system then the "out-the-box" alert... I am currently looking into the SWQL but not having much luck.
The environment has lots of AP's and most are connected to more than 1 controller (Sometimes up to 3) and here is the full scenario of what I am looking for:
Currently setup (Out-the-box)
When an AP goes down - Get an Alert
Example: AP 10.20.30.10 is connected to Controller 1, Controller 2 and Controller 3.
When the node goes down on each controller, I will get an alert per controller.
NEW Alert 1 - Thin AP fully Down
This alert should only be triggered when the thin AP is classed as down on all connected controllers.
Example: AP 10.20.30.10 is connected to Controller 1, Controller 2 and Controller 3.
The Following condition for this AP has to be met "Down on all 3 controllers"
IMPORTANT NOTE: Not all of the AP's on the network are the same (Some only have 1 or 2 controllers attached), so I would believe there would need to be some if/and logic statement like below:
* Total Controller count = Where AP Name x Related Node
if
AP Name = Down
And
Related Node count = Total Controller count
IF the condition above is met, then raise the alert that the Thin AP point is down
NEW Alert 2 - Thin AP "Leg Down"
Like how alert 1 is structured, if the Thin AP is not "Fully down" then it is only partly down because it has lost contact to one of the controllers that is not equal to maximum.
Example: AP 10.20.30.10 is connected to Controller 1, Controller 2 and Controller 3.
The Following condition has to be met "Down on 1 or more controllers, but not equal to the total number of connected controllers"
* Total Controller count = Where AP Name x Related Node
if
AP Name = Down
And
Related Node count < Total Controller count
IF that condition above is met, then raise the alert "THIN AP leg down"
Hope I have provided you Thwack Brainiac's both a challenge,and something that can be achieved
