So kind of an open ended question here - How do you have your health monitors setup? Do you just monitor the web server, do you have something setup to monitor the backend database?
In a much simpler time our web applications had SQL strings in the web.config file but also had a dependency on UDL files. Our health monitor is a file in the root of the site that references that UDL as a connection string and queries the database, so the health monitor is looking for a specific response. This way we are simultaneously testing the web server and the database. We no longer use UDL files on our product (with the exception of legacy stuff still lingering), plus its caused us to build a dependency on a database that a specific product may not even use. Plus side though is that staging and production nodes can go in to the same pool and we can toggle back and forth if need be based on that file (we use different file names for staging than we do in production). We also have a less than reliable web admin that is frequently having issues with this system.
So Im looking for an updated way to do my health monitoring that monitors the web server and database preferably. How are you guys doing it? Any novel ways of monitoring the application stack?
You could create a page called test.html that will perform a database query. If the database is up and running (and the webserver to) it will give a response text of "everything is ok". If the database is down it could give another response.
In the health monitor you can modify the Receive String value to everything is ok. This will give the poolmembers the Available status. Any other response will mark it as down.