parca VS client_golang

Continuous profiling tools such as parca may be worth looking into for your use case.

They were bought by Elastic, maybe they'll open source it. There's also https://github.com/parca-dev/parca

an alternative option to pyroscope to do continuos profiling in production could be parca.dev check and here

+1 to Parca.dev https://github.com/parca-dev/parca as continuos profiling tool in production

Parca collects, stores and makes profiles available to be queried over time. It is open source and can be deployed on production environments as Parca focuses on sampling profiling two main types of profiles: tracing and sampling.

Polar signals develops Parca [0] which is another eBPF observability tool, and Isovalent develops Cilium [1] which is built on eBPF as well. Genuinely curious if there are differences, or if eBPF only allows for specific observability functionality and each tool has it all.

[0]: https://github.com/parca-dev/parca

[1]: https://github.com/cilium/cilium

Prometheus, an open-source systems monitoring and alerting toolkit, has client libraries that allow you to instrument your services in a variety of languages. Here's an example of how you might use the Go client library to expose metrics on an HTTP endpoint. We will leverage the prometheus go client to expose metrics and create them.

Monitoring prometheus, grafana

I'd recommend github.com/victoriametrics/metrics for exporting custom metrics in Prometheus format. It is easier to use than the github.com/prometheus/client_golang.

Start at the official documentation and the Go client library .

DataDog can read Prometheus metrics. Note that this approach works for metrics but not for traces although you can configure DataDog to receive OpenTelemetry traces as well https://github.com/prometheus/client_golang

For Go, the standard is to use Prometheus and the Prometheus Go client library.