Cloud-Native Observability @ BASTA Spring 2023

Moderne verteilte Anwendungen sind wie viele kleine Zahnräder, die perfekt ineinandergreifen sollten. Dabei lässt es sich in der Praxis meist nicht vermeiden, dass es doch hier mal knirscht oder es dort mal kleine Reibungsverluste gibt. Aber nur, wenn unsere Cloud-Native Anwendung auch sinnvoll überwachbar ist, können wir solche Problemstellen erkennen, identifizieren und beheben. Und zwar idealerweise, bevor es sich derart verklemmt, dass die ganze Anwendung stehen bleibt.

Thorsten Hans zeigt Ihnen in dieser Session, wie man mit Standard-Vorgehensweisen, -Tools und -Protokollen der Cloud-Native-Welt (Stichwort Observability mit OpenTelemetry) Logs, Metriken und Traces in der eigenen Anwendung erheben kann. Zudem erleben Sie, wie man diese Informationen dann auch sinnvoll anwendungsübergreifend zusammenfassen und auswerten kann, sodass potenzielle Probleme schon frühzeitig auffallen können.

Transcript

1. Cloud-Native Observability Thorsten Hans Cloud-Native Consultant @ Thinktecture
3. Hi 👋🏼, I'm Thorsten Hans Cloud-Native Consultant @ Thinktecture Technology Focus #CloudNative #Kubernetes #Azure #Go #Containers #Rust #WebAssembly Awards Microsoft MVP for Azure Docker Captain Contact thorsten-hans.com thinktecture.com thorsten.hans@thinktecture.com | @ThorstenHans
4. Talking Points What is Observability Observability in Cloud-Native OpenTelemetry Instrumenting a .NET app Alerting Conclusion
5. SECTION What is Observability
6. What is Observability With Observability we gain insight into complex systems by monitoring their states and behaviors It involves collecting and analyzing data from various sources Observability helps to identify and diagnose issues in systems We use logging, metrics collection, and distributed tracing Observability allows developers to better understand how their systems are performing in real-world environments
7. How Observability looks like Distributed Traces
8. How Observability looks like Metrics
9. How Observability looks like Logs
10. DEMO Let's explore traces, metrics, and logs
11. SECTION Observability in Cloud-Native
12. Observability in Cloud-Native Cloud-native applications are highly distributed, complex systems that can be difficult to monitor without observability Observability provides a way to gain insight into the behavior of cloud- native apps By implementing observability practices, developers can collect and analyze data from various sources to gain a comprehensive view of their cloud-native app. Especially in cloud-native apps, it makes performance penalties discoverable
13. Common Observability Tools Distributed Tracing Jaeger Zipkin Grafana Tempo Metrics Prometheus Grafana Metrics Logs Grafana Loki Elastic Search Log Shipping FluentBit (or FluentD) Logstash Promtail Visualization Grafana Kibana
14. SECTION OpenTelemetry
15. OpenTelemetry OpenTelemetry (otel) https://opentelemetry.io/ OpenTelemetry is a set of open-source libraries for collecting and analyzing telemetry data from cloud-native applications It is available for most popular programming languages It works with a wide range of cloud environments and platforms It provides an agnostic standard for telemetry data collection It makes it easier for developers to instrument their apps
16. OpenTelemetry in .NET OpenTelemetry is a collection of pluggable NuGet packages of which we’ve to pick and choose: # Fundamental packages dotnet add package OpenTelemetry dotnet add package OpenTelemetry.Extensions.Hosting # pick and choose intrumentation packages dotnet add package OpenTelemetry.Instrumentation.Runtime dotnet add package OpenTelemetry.Instrumentation.AspNetCore # pick and choose exporter packages dotnet add package OpenTelemetry.Exporter.Console dotnet add package OpenTelemetry.Exporter.Zipkin
17. SECTION Hands-On: Instrumenting a .NET app
18. DEMO Instrumenting a .NET app with Open Telemetry
19. SECTION Alerting
20. Alerting Notification engines (e.g. Alertmanager) inform about anomalies based on configurable condiditions Alerts are generated by systems like Prometheues Alertmanager consolidates alerts and actually notifies people Common messaging patterns are built-in: E-Mail Pager Chat messages (e.g. Teams | Slack)
21. DEMO Alerts for custom metrics
22. SECTION Conclusion
23. Conclusion Observability is critical for maintaining the reliability and performance of cloud-native apps. Observability makes the health of an app visible and actionable. Custom metrics are key to identify KPIs OpenTelemetry (otel) is defacto-standard for instrumenting apps it intergates with self-hosted and SaaS targets
24. Thank you @ThorstenHans | thorsten.hans@thinktecture.com | GitHub Repository 👋🏼 Thanks for attending my talk 👋🏼