Prometheus Alertmanager Configuration

[Ameni Elheni]

Hi, I'm new here and need your help. I'm working on my graduation project and one of my tasks is configuring the Alertmanager. Unfortunately, I couldn't achieve that. I'm working with AKS and installed the Alertmanager using the helm chart, stable/Prometheus-operator. I tried to change my values.yaml file but it didnt work out! Any help? I really need it. 

Thank you in advance. 

[Brian Candler]

Not from that description I'm afraid.

If you have a *specific* question then please ask it.  e.g. you used a given alertmanager.yml, you started it, and (a) it failed to start - if so what was the error message; or (b) it started but didn't behave the way you expected - in which case, what did you do, what did you expect to see, what did you see instead.

Also you should note a couple of things:

- the prometheus operator is not part of the prometheus project.  Problems with the operator should be asked on a forum related to that project.

- the prometheus operator is now obsolete anyway.  See the clear deprecation notice at https://github.com/helm/charts/tree/master/stable/prometheus-operator which points you to the newer project you should be using instead.

[Ameni Elheni]

  Thank you for the notes. My query is how to configure my alertmanager. I couldn't find the alertmanager.yml file. I only had the values.yml file, so I tried editing it by adding the email notification configuration, but then my alertmanager didn't work at all. The pod's status was "CrashLoopBackoFF". Should I create it on my own? I don't know what to do. All I want is to get the alerts and configure the email/slak notifications part...  

[Brian Candler]

"CrashLoopBackoff" is a kubernetes status.  Look at the logs from the pod (kubectl logs ...) to see the error message.

If you can't find alertmanager.yml, it's because you're not using alertmanager directly but deploying it via an operator.  As I say, this isn't a problem with alertmanager, it's a problem with the operator.  The operator isn't part of prometheus, so you'll need to ask the operator project for help.  Or you could run alertmanager yourself.