Parity of metrics between Atlas UI and Prometheus integration
It would be great if any metrics that are available in the Atlas UI are also available via the the Prometheus integration, especially since the metrics in the Atlas UI get coarse-grained after a while.
For example, in a recent incident it took us a while to do some preliminary investigations, and we wanted to confirm by looking at historical values for the metric "Operation Execution Time".
Unfortunately the chart in the Atlas UI was already too coarse-grained and the metric also does not seem to get exposed via the Prometheus integration.
9
votes
Johannes Pirkl
shared this idea
-
Gregory
commented
`rate(mongodb_opLatencies_reads_latency{ instance="cluster0-shard-00-02.v9uas.mongodb.net:27018"}[5m])/rate(mongodb_opLatencies_reads_ops[5m])/1000`
works well for me, you can adjust the time granularity