Take me Down to the Paradise City Where the Metric is Green and Traces are Pretty
A presentation at SRECon 2021 by Ricardo Ferreira
Take me Down to the Paradise City where the Metric is Green and Traces are Pretty Ricardo Ferreira Principal Developer Advocate, Elastic
Who am I? • Principal Developer 🥑 at Elastic • Community, Developer Relations • Before Elastic: Confluent, Oracle, Red Hat • Observability, Streaming Systems, Databases, 🧑💻 Java and Golang • https://riferrei.com @riferrei
Agenda • Observability 101, redefined • Overview about OpenTelemetry • Creating an observability backend • Example of instrumented app in Java • Example of instrumented app in Go
Current perception about observability Where When What Tools + Processes + People Tools + Processes + People Tools + Processes + People Logging Metrics Tracing
“…this seems like calling ‘gasoline, motor oil, and tires’ the 3 pillars of F1 racing. It is not wrong, precisely, but…” Josh Berkus Kubernetes Community Manager, Red Hat
Observability redefined: pipes (not pillars) Request-scoped metrics Metrics Aggregatable Tracing Aggregatable events Request Scoped Logging Events Request-scoped events Request-scoped contextual View
Overview about OpenTelemetry https://opentelemetry.io
• Based on the projects OpenTracing and OpenCensus • Open source CNCF project • Provides specification, protocols, SDK, and integration tools • Unifying all types of telemetry data: traces, metrics, and logs • Vibrant growing community
Different deployment topologies
Is OpenTelemetry stable enough?
• Traces are GA and programming OpenTelemetry Stability languages are mostly up to speed • Metrics are almost GA and programming languages are mostly up to speed • Instrumentation libraries are popping up every single week. Help to accelerate this. • Logs are not ready. But you can use the collector for most things log-related. • TL;DR: yes… start using OpenTelemetry
Time to see some hands-on stuff
Creating a observability backend Microservice OTel Collector Microservice ?
Creating a observability backend Microservice OTel Collector Microservice Elastic Observability
Get started with Elastic for free today! Try now: cloud.elastic.co/registration
Example of instrumented app in Java Microservice OTel Collector Elastic Observability https://github.com/riferrei/otel-with-java
Example of instrumented app in Go Microservice OTel Collector Elastic Observability https://github.com/riferrei/otel-with-golang
How to find me? Follow me: @riferrei YouTube Channel: videos.riferrei.com Previous talks: talks.riferrei.com
Thank You
Observability is a software discipline that goes back to when virtually any problem could be solved by tailing web server logs. But the world has changed, and systems these days are comprised of different services running in their own stacks who cooperatively build up what we understand as the end-to-end architecture. Thus, observability had to evolve as well.
Today we have OpenTelemetry—an observability framework for cloud-native software. OpenTelemetry provides the tools, APIs, and SDKs to create a reusable, robust, and non-vendor-driven observability strategy. But the reality is that most developers are still confused about the lines that separate OpenTelemetry from the past and which parts of the framework are stable enough to be used in production. This talk will explain how OpenTelemetry works and provide examples in Java and Go to illustrate the APIs you can use to produce traces and metrics.
Video
Code
The following code examples from the presentation can be tried out live.
Buzz and feedback
Here’s what was said about this presentation on social media.
