Ok, here is my scenario.
I have a highly available site, that has about 20 different subsites as critical components. These are all behind a load balancer. We have been asked to do some alerting on these, beyond simple http(s) site up/down. For auditing purposes we want the ability to monitor if site X is unavailable on ALL of the backend servers, and ONLY alert in that case. Including the name of the IIS site in the alert itself as a passed variable.
I think i am going to have to build 20 individual alerts here, and i am looking for a more intelligent way to do this.
so from a high level it would look like this.
For each site
On server 1
on server 2
on server 3
is the site up?
if its not up on all 3, that's my trigger condition.
if its up on only 1 of them, no trigger.
my problem is that if i list all 3 servers as an AND, that condition seems to translate as the node name must equal all 3 names.
if they are listed as an OR, it trips for any of them.
Is there a good way to perform a for-each style alert from the standard alert generator using the application insights for IIS :Sites monitor?
