Panel Discussion: Application Monitoring
summary
Panelists compare tracing, profiling and logging with real examples: prioritize profiling for typical wins, use tracing for deep dives, and rely on sampling to control cost; Node apps break when running multiple APMs due to monkey patching conflicts. They name Datadog and InfluxData, recommend Slack/PagerDuty alerts, and cite compliance needs like HIPAA and credit card monitoring catching fraud. Future work emphasizes eBPF and AI for targeted heavy tracing and predictive maintenance, illustrated by a metric drop from 10,000 to zero clicks.
AI-written summary · about
chapters (14)
- 0:00Panel introduction
- 1:03Defining application monitoring and its importance
- 3:50Advice for new engineers
- 6:18Common monitoring mistakes
- 9:06Monitoring for different teams
- 11:33JavaScript metrics to watch
- 14:21Monitoring for security
- 18:17Future of monitoring
- 23:11Final monitoring advice
- 24:42Effective alerting strategies
- 28:33Alerting tools evolution
- 31:45AI in performance monitoring
- 33:08Challenges and benefits of using multiple APMs
- 35:20Towards observability
related talks
50mPanel: Observability
33m100% Visibility
19mOops my app performance sucks... what if observability could be a game changer?
58mObserving Applications in Production
24mFull Observability for Application Monitoring
52mHolistic Monitoring
23mMastering Monitoring & Alerting: Key to Seamless Performance
33mWhat Developers Should Know About Observability