Instrumenting your Java applications with OpenTelemetry

Incubating for three years, OpenTelemetry is already the 2nd most active CNCF (Cloud Native Foundation) project behind Kubernetes, and became a de-facto standard for observability. This framework allows instrumenting applications, generating, collecting and exporting telemetry data such as traces, logs and metrics without vendor-locking.

In this talk, we will discuss and code the different ways to instrument your Java applications using OpenTelemetry tools such as its API / SDK, other frameworks support, and agent-based automatic instrumentations. By the end, you should be able to generate telemetry data and start contributing to your system observability.

Transcript

1. Instrumenting your Java applications with OpenTelemetry June 2023
2. Bruce Bujon @HardCoding Software Engineer, Datadog Paris JUG Leader AdAway Developer
3. The OpenTelemetry Realm 3 Images generated with Craiyon
4. The OpenTelemetry Realm 4
5. The OpenTelemetry Realm 5
6. The 3 Observability Pillars 6 Logs Traces Metrics
7. The 3 Observability Pillars Traces only ⏱ 7 Logs Traces Metrics
8. 8 Demo
9. Instrumentation Options Summary 9 Framework Support • Good generic instrumentation • Only custom instrumentation needed • Support and configuration varies per framework • Limited control over signals OTel auto-instrumentation • The most complete instrumentation • The best support of libraries • Only custom instrumentation needed • Configuration vendor specific • JVM agent required OTel API/SDK • Full control of instrumentation • Good knowledge of OTel required • A lot of custom developments • Painful to instrument libraries
10. 10 Starting with OpenTelemetry is only 1 line away
11. Thank you Questions? Bruce Bujon @HardCoding