Monitoring your platform from multiple locations
summary
Claim: tooling is easy but the hard part is design so the talk shows concrete choices not platitudes — using Postman, DataDog, New Relic, AWS, Uptrends, Cloudfare as examples, selecting a mix of countries and regions, running monitors once per minute (about 45,000 entries per month ~half a million per year), flagging probes via allowlisting and headers, retrying before alerting, grouping alerts to avoid storms, and a working Typeform setup that proved the approach and iterated from defaults to custom thresholds.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:17Scaling global platforms
- 3:43Multi-region monitoring improvements
- 4:49Monitoring strategy questions
- 6:37Equal alerting across locations
- 7:19Monitoring frequency and granularity
- 8:23Security for monitoring
- 8:44Alerting strategies
- 9:45Failure criteria definition
- 10:47Incident response teams
- 11:53Redundant alerts handling
- 12:55What to monitor
- 14:43Conclusion
related talks
44mGo global, or go home. Monitoring your platform from multiple locations
23mMastering Monitoring & Alerting: Key to Seamless Performance
32mCustomer Centric Monitoring
27mBuilding a Scalable Monitoring System
52mHolistic Monitoring
58mObserving Applications in Production
37mIntroducing Reliability Toolkit: Easy-to-Use Monitoring and Alerting
39mPanel Discussion: Application Monitoring