Open ankitnayan opened 3 months ago
We should derive a graph of connected components and show the relevant metrics of the dependent components (and resources that these components are part of). When the alarm goes off for any metric, it's usually because of the upstream/downstream and the resources these components rely on. We should highlight the hotspots of the issue instead of the user trying to hop from one dashboard to another to figure out what is wrong.
Today traces and logs explorer links are part of alerts but not metrics. The alert_id which is part of fired alerts can take you to the alerts page to view charts during that interval but the actual filters are not populated automatically which were part of the alert.
It is difficult to analyse data from metrics alerts. What more use cases can be solved apart from filters being automatically applied when going from alerts to metrics explorer page?