Observability in strikethrough action! /strikethrough the wild
summary
A cloud backend on OpenShift (Azure Germany) runs ~70 microservices in production and ~120 in staging scaling to 200–300 pods, using Prometheus + Grafana for metrics, Zipkin for tracing, Elasticsearch/Kibana and Fluentd for logs. A chatbot exposes a verbose flag links trace and logs to jump from a user request to Zipkin spans and log queries; tracing stays off by default but can be activated per session with per-user tracing via access token. Result: faster bug reproduction, testers include trace IDs, first-level support can diagnose without experts.
AI-written summary · about
chapters (8)
- 0:00Introduction and speakers
- 4:30Collaborative monitoring vision
- 7:28Current monitoring toolchain
- 11:07Chat bot integration demo
- 18:55End-to-end test integration
- 21:37High-level setup overview
- 23:26Summary and lessons learned
- 28:46Challenges and acceptance
related talks
50mDistributed Tracing in the Wild
24mDistributed Tracing : New DevOps Foundation.
01h 02mAdding Observability to Distributed Systems
33mOperational and Business Insights from Tracing Distributed Microservices
42mDistributed Monitoring: How to understand the Chaos
53mUnmask hidden issues with Observability superpowers!
52mIs your JVM app flying blind? Unmask hidden issues with Observability superpowers!
01h 07m"Adding Observability to Distributed Systems"