Where do I configure SMTP

[mhabicht]

I have been unable to determine where I configure SMTP. I have checked all the yaml and config files in /prometheus-operator and reviewed the value.ini in /prometheus-operator/helm/grafana. Am I supposed to create the grafana.ini or is there another file this configuration gets added to?
My Alert manager pod is running and sendmail is configured, but I still get an error when I try to send a test message that SMTP is not configured in grafana.ini.

Thanks

[brancz]

If you are using the helm charts of the prometheus operator repository then this repository has nothing to do with what you are deploying, this repository is only relevant if you are using jsonnet to render the manifests.

[mhabicht]

[brancz]

Are you sure you are using the jsonnet based stack? You mentioned helm in the original issue. This is not helm.

[mhabicht]

The mention of helm was the directory location. The original install was "prometheus-operator" When I posted there I was given a link and told to post here. prometheus-operator/prometheus-operator#1972.
I see the custom-ini.jsonnet in examples, what directory do I add that to?

[brancz]

If you are using helm, then you cannot use that, you must use the jsonnet based stack: https://github.com/coreos/prometheus-operator/tree/master/contrib/kube-prometheus

[mhabicht]

I have these 2 files, prometheus-operator.jsonnet prometheus-operator-rbac.jsonnet, neither appear to have the correct context for SMTP.

[brancz]

Those are unrelated. I encourage you to go through the readme in the link I posted above.

[metalmatze]

Once you have set these paths inside the Grafana config you need to mount those, correct.
You need to merge the Grafana object we ship with extra configuration that you need, more precisely the secrets to mount.

Openshift's cluster-monitoring-operator is a good example:

You already have the configuration extended:
https://github.com/openshift/cluster-monitoring-operator/blob/1cff58efb310c00d67ea48135136fc2e2242d945/jsonnet/grafana.jsonnet#L49

Want you need is to extend the deployment itself:
https://github.com/openshift/cluster-monitoring-operator/blob/1cff58efb310c00d67ea48135136fc2e2242d945/jsonnet/grafana.jsonnet#L188