Prometheus and "No metrics store"

[tacomablues]

I may be reading something wrong, but I consistently get a "No metrics store has been defined" error message when logging into the admin portal of my Stalwart instance. First, let me confirm that I do have an Enterprise license for my Stalwart installation. I have installed Prometheus on the Stalwart server and have things properly configured where Prometheus is successfully scraping the Stalwart Prometheus endpoint. I can confirm this by logging into the Prometheus web interface and checking Status -> Target Health. I see the Stalwart endpoint with an up status.

I was under the impression that the metrics exported from Stalwart into Prometheus could then be read back into Stalwart to populate the "Messages" and "Memory Used" graphs under the Stalwart admin portal dashboard. The catch is I don't know where to find any such settings, if my assumption is accurate. The only reference to Prometheus metrics that I can find in the Stalwart admin port is under Settings -> Telemetry -> Metrics -> Prometheus Pull Metrics. I have the Username and Secret fields blank and the "Enable endpoint" enabled. These three settings are simply used to configure the export of metrics from Stalwart for the Prometheus scraping process.

In short, am I mistaken in my above assumptions? If not, how does one pull the Prometheus metrics back into Stalwart to populate the given dashboard graphs? Or can someone point me in a completely different direction on how to populate these dashboards graphs? My thanks for any feedback on this topic.

[tacomablues]

Okay. Seems I answered my own question. I had to go Settings -> Telemetry -> History. There I found the option to set a Postgresql store as this is the database backend I have for my Stalwart installation. I enabled history for both the tracing and metrics stores. After this I restarted Stalwart and found the metrics store error message was removed. Seems that the dashboard graphs are populated using the local database. Prometheus is only for external metrics reporting via Grafana, et al. Thank you.