#Best APM for nest/node?

8 messages · Page 1 of 1 (latest)

ornate bough
#

Any recommendations for some sort of application performance monitor for Nest? I'm trying to track down some high CPU usage (culprit looks like TypeORM) but I need more insight as to what each request is doing: I need to break it down by timing functions, DB calls, etc.
if it matters, this is in a fargate/containerised environment. Thanks in advance!

unique gorge
#

Depends on the stack. Express, Apollo GraphQL and other defaults are usually well supported. Fastify, mercurius, not so much.
I prefer using Sentry Tracing, had some luck with Newrelic, and am trying to avoid OTEL (open telemetry), but otel is probably your best bet.

ornate bough
unique gorge
#

Yeah, no official fastify integration. I had the same struggle, wasn't satisfied with Newrelic/Datadog, until I eventually wrote my own little integration for Mercurius. It's not that difficult for basic tracing, but its going to be difficult trying to tie the requests to "what happens during them" (ie the span tree view/breakdown).

Also found https://autometrics.dev/ recently, but haven't tried that one yet.

Autometrics is a developer-first observability framework for understanding the error rate, response time, and production usage of any function in your code. Implemented in Rust, Typescript, Golang and Python. Open Telemetry and Prometheus compatible.

bright palm
#

u can check also insnight performance info in aws 🙂

nimble shoal
#

I would say go with OpenTelemetry (and the community integration nestjs-otel) Most other tools are built on top of that anyway

stray elbow
#

I love honeycomb, easy to setup, really usefull metrics

lethal imp
#

I work with datadog in my company and we are pretty happy with it