Trickster currently works as a proxy sitting between Grafana and a TSDB Origin (currently just Prometheus). It would be really cool if we could also modify Trickster so it can consider Grafana an origin too. That way, a single Trickster will adapt to any new data sources added to Grafana, without having to update the Trickster config to have specific knowledge of them. This only matters for Grafana-proxied data sources (non-proxied will work as normal and will hopefully use their own Trickster instances). In this hypothetical Grafana origin mode, Trickster will need to keep tabs on which Grafana proxy path maps to which origin datasource type. So it would need to know how to ask Grafana for this information, both on startup and any time a new/unrecognized proxy path is requested via Trickster.
Dependent on #9 completion
Trickster currently works as a proxy sitting between Grafana and a TSDB Origin (currently just Prometheus). It would be really cool if we could also modify Trickster so it can consider Grafana an origin too. That way, a single Trickster will adapt to any new data sources added to Grafana, without having to update the Trickster config to have specific knowledge of them. This only matters for Grafana-proxied data sources (non-proxied will work as normal and will hopefully use their own Trickster instances). In this hypothetical Grafana origin mode, Trickster will need to keep tabs on which Grafana proxy path maps to which origin datasource type. So it would need to know how to ask Grafana for this information, both on startup and any time a new/unrecognized proxy path is requested via Trickster.