Monitoring Kubernetes VS Serverless-based applications
summary
Compares Kubernetes-era monitoring (Prometheus + Grafana, ELK, service mesh like Istio) with serverless realities and claims agentless monitoring is mandatory because functions are ephemeral and agents cannot be deployed. Argues for automatic distributed tracing across managed services and shows a demo where a Lambda runs ~7,000 times with ~50 failures, Lumigo stitches metrics logs and traces into an end to end request story across EventBridge DynamoDB Twilio with no code changes and quick setup.
AI-written summary · about
chapters (10)
- 0:00Monitoring Kubernetes vs serverless
- 1:02Cloud computing evolution
- 2:29Kubernetes monitoring layers
- 3:30Kubernetes monitoring tools
- 4:56Service mesh for monitoring
- 5:37Serverless environments
- 6:20Monitoring serverless differences
- 7:43Modern cloud monitoring needs
- 8:25Serverless monitoring demo
- 12:19Key takeaways and challenges
related talks
28mWhat We Should All Worry About When Monitoring Serverless Applications
41mIs Serverless the End of Kubernetes?
05mA Serverless Monitoring Checklist
35mWhat does Serverless mean for tomorrow’s abstracted infrastructure?
18mObservability in Serverless Applications
49mMonitoring serverless applications on AWS: the monitoring monster becomes your friend.
34mObservability in Serverless Application
59mModernize faster with serverless and end-to-end observability