Alerting "no such host" error when running Prometheus via reverse proxy

Hi there,

I have a setup where Grafana (v5.2.4) and Prometheus (v2.3.2) are on Kubernetes and are both running behind Traefik under separate domains: e.g. grafana.minikube and prometheus.minikube. (they’re both declared in my /etc/hosts file). They can both be accessed fine through the browser. For the datasource I had to select “Browser”, because with ‘Server’ I was getting a bad gateway error, when testing the Prometheus datasource. Unfortunately, alerting doesn’t seem to work, as I’m getting “prometheus.minikube no such host” error, when testing an alert. Any idea of what might be wrong ? it seems to me that Grafana server can’t access prometheus API due to the domain declaration, even though it can be accessed via the browser. Anything that needs to be declared via environment variable ?

Did you find answer on your question?