Improve Observability with OpenTelemetry

Issue: Improve Observability with OpenTelemetry

Our application currently lacks proper observability capabilities. We would like to implement OpenTelemetry to collect metrics, traces, and logs from our application and services.

One question is whether we need to manually instrument our code to generate these telemetry data points, or if we can leverage auto-instrumentation techniques. eBPF is an option for auto-instrumentation that looks promising but requires further research into its capabilities and performance characteristics for our use case.

To address this, create a task to research eBPF and evaluate its suitability for auto-instrumenting our application to collect OpenTelemetry data without code changes. The research should focus on understanding what types of instrumentation are possible with eBPF, any limitations, and potential performance impacts of using it for monitoring versus manual instrumentation. The findings will help inform our decision on the best approach for OpenTelemetry implementation.

---

Description was generated using AI

Prompt:

Our application needs better Observability. We want to evaluate metrics, traces, and logs using OpenTelemetry. The challenge is whether we need code instrumentation, or can use auto-instrumentation. Add a task to research on eBPF for auto-instrumentation, and performance profiling.