Monitoring Microservices — Part 2: Using SLOs to guide Monitoring Objectives

Monitoring with the end customer in mind

Photo by Guillaume Bourdages on Unsplash

In the previous article, we unveiled the curtain on microservices observability. Now, I invite you to venture into instrumentation, analytics, and notifications, all steered by Service Level Objectives (SLOs). By adopting SLOs, we shift our monitoring perspective to prioritize the end-user experience.

NOTE: Because it took so much time to write part 2, I decided to write part 2 including some tips (or best practices, if you prefer 😅).

Instrumentation: The Data Harvest

Instrumentation is our lens into the system’s performance from the user’s standpoint.

Instrumentation is our gateway to the microservices’ heartbeats. It’s all about embedding sensors in the code to capture the performance rhythm and system behavior.

• Tracing: Trace requests across microservices to measure latency against SLOs like “99% of requests complete within 300ms” (Jaeger, Zipkin).
• Metrics: A metrics-driven approach keeps you in the know about system vitals like response times, error rates, and more. Measure these system vitals and compare them…