Spend more time innovating, less time troubleshooting
summary
Demonstrates Dynatrace running an agent that collects JVM .NET and JavaScript RUM plus logs and cloud metrics, builds a live dependency model, and uses machine learning to surface root causes not raw alerts. Example: a real incident shows 61 users and ~700 service calls affected over 27 minutes, 38 individual anomalies correlated into one correlated "problems" from 38 anomalies, with the failing process auto‑restarted in ~5 minutes but full user recovery taking ~30 minutes. Scales to ~15–20k hosts as used by SAP.
AI-written summary · about
chapters (22)
- 0:00Introduction and agenda
- 0:20Why traditional monitoring fails
- 1:03Digital transformation challenges
- 2:31Limitations of classic monitoring
- 3:33Complexity of modern applications
- 4:57Software as competitive edge
- 6:00Changing environments and monitoring
- 7:22Social media shitstorms impact
- 9:09Machine learning in monitoring
- 10:35Automating monitoring with AI
- 11:37Data collection and modeling
- 13:23Live product demonstration
- 14:27Monitoring dashboard overview
- 15:33Interactive system model
- 16:39Infrastructure monitoring details
- 17:25Application performance insights
- 18:07Anomaly detection and learning
- 19:12Correlated problem analysis
- 21:44Scaling monitoring with AI
- 22:27Root cause analysis demo
- 23:55Value proposition summary
- 24:16Closing and next steps
related talks
46mTroubleshooting Distributed Workloads ...
01h 00mData is out, knowing is in: Applying AI to automate cloud operations
15mIntelligent Incident Management in DevOps
20mSuper vision smart goggles
34mIncidents Happen
47mBridging runtime and build time intelligence to reduce friction
32mCustomer Centric Observability: How Intuit Reduced Time To Detect Customer Impact From 30+ Minutes To Under 3 Minutes.
01h 05mThe art and methodology of troubleshooting