Hi, we need some help getting Dynatrace to alert properly. We are doing preliminary tests to ensure that our monitoring and alerting is working prior to launch. One of the most basic things we want to ensure is that if there are no ECS Fargate tasks to accept our API calls, we get alerted on it. In order to test this, we simply shutdown all tasks on our cluster and validated that they were down by hitting the health check url and got a 503.
We left them down for 15+ minutes. During this time, we can see the tasks on the hosts screen go to an unmonitored state but no problem ever gets generated. Can someone let me know what setting we need to enable to alert in a scenario where there are no tasks running for a particular ECS cluster?
So I'm just trying to think with you.
Do you have this service monitored? Do you see calls?
I want to understand firstly if this service is monitored and you see to healthy calls,
As much as I know, in order to see low load / failure rate (because of the 503) the service first build a baseline to understand what is normal and on what to alert.
We are still in development so there is no baseline yet. Here's what I need to do.... There should never be a case where there are 0 tasks for this Fargate based api. If there are 0 tasks we want an alert. I do not want to have to build a synthetic for this. I want Dynatrace to alert me if there are no tasks to support this api.