gnort is a Datadog client library that provides efficient in-process metrics aggregation. I wrote this because I wanted to be able to do a mixture of aggregated and ad-hoc metrics with minimal boilerplate. Accordingly, a lot of the value of this library is the codegen via macros, registry, and aggregation windows.
I say this a "Datadog" library because the aggregation windows and push-based mechanisms aren't really compatible with the assumptions of the Extended Prometheus Cinematic Universe. I find Prometheus makes metrical analysis and processing more difficult rather than easier although I do appreciate why they went with pull-based metrics.
This library does use the dogstatsd crate under the hood, but I could imagine making it more generally "statsd" oriented.
I'd like it if Datadog made distributions something we could aggregate client-side. Daddy needs his count-min sketch.