Using Open Source Observability with Lambda
summary
Shows how to replace CloudWatch blind spots by shipping Lambda logs, metrics and traces into open source stacks: ELK (Elasticsearch, Logstash, Kibana) for logs via a CloudWatch-to-Kibana shipper lambda, Prometheus + Grafana for metrics using Metricbeat pulling CloudWatch namespaces, and Jaeger tracing (containerized) for spans. Builds a Serverless Framework demo that polls Giphy and post-processes images, runs Metricbeat in Docker and a log-shipping Lambda; the pipeline runs end-to-end (some demo errors expected), proving the approach is practical.
AI-written summary · about
chapters (8)
- 0:00Intro
- 1:08What's the problem?
- 3:37The next step...observability?
- 5:18Event logs
- 8:44System metrics
- 11:34Tracing
- 15:23Demo
- 27:24OpenTelemetry
related talks
34mObservability in Serverless Application
41mLive Coding on Observability for Serverless
43mObservability for Serverless
32mBuilding and scaling a log analytics platform - a serverless approach
24mObservability across serverless asynchronous managed services
30mUse the Observability, Luke!
39mObservability in DevOps
01h 02mDesigning an open-source observability stack for Kubernetes