I am deploying Grafana via the kube-prometheus-stack Helm Chart. I initially configured the number of replicas (global.replicas) to 1. Everything works find and I can successfully render the dashboards on Grafana. However, once I increase the number of replicas to, let’s say 3, The Dashboards are not displayed and I got in the logs the following error:

[guardian.dashboardNotFound] failed to get dashboard by UID grafana

If I port forward to any of these Grafana replicas, I can render the dashboards properly. What could be the problem? Am I missing something here in the configuration?

Any hint would be appreciate it.
Thanks!