Export profiler slow queries via Datadog integration
The "profiler" tab in Atlas is very nice -- it gives a really nice over-all view of slow queries, with the ability to drill down into the individual calls.
However, I don't check this page every day, and so it is sometimes hard to find out when we actually do have slow queries or commands pop up on it if it is not affecting our cluster.
We use datadog as the central point for all of our monitoring. We have the Atlas <--> Datadog integration set up, which exports a number of useful stats. It would be very nice for the profiler results to be auto-exported to datadog via the integration, because then we could:
- configure datadog to keep them for as long as we want
- configure alerts and/or dashboards on this data as we see fit
I do know that one work-around for this is to manually wire up something to query the profiler REST endpoints and push that to datadog. This is likely what I will end up doing now, but it would make things a lot easier if I (and all other customers using this integration) got this data in datadog "for free" (free in terms of effort).