Serverless Architecture Patterns and Best Practices
summary
Practical advice on building serverless apps with AWS Lambda: use AWS SAM to cut a CloudFormation template from ~80 lines to ~20 and test locally with SAM local (Docker), build CI/CD with AWS CodePipeline + CodeBuild and deploy via CloudFormation, measure with CloudWatch (log pivots, percentiles) and trace with AWS X‑Ray, use parameter store for secrets, adopt weighted deploys via traffic shifting in lambda aliases (example 5% canary), expect cold starts under 0.25% and often reduce cost by increasing memory/CPU.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 1:08What is Serverless
- 8:36Where to Start
- 11:16WSSam
- 14:59Code Build
- 18:49Stage Variables
- 27:17CloudWatch
- 28:45Xray
- 30:47Troubleshooting
- 33:22Best Practices
- 33:46A horrible idea
- 33:56Lambda functions
- 34:33Languagespecific packages
- 35:52Lambda performance
- 36:26Memory vs CPU
- 37:37Cold Starts
- 39:14Separating Business Logic from Function Signature
- 40:02Connecting to Global Scope
- 40:37Virtual Private Cloud
- 42:29CloudWatch Hack
- 44:10Wrap Up
related talks
49mServerless Architectural Patterns and Best Practices
50mServerless Architecture Patterns and Best Practices
04mServerless applications, Microservices, AWS
42mNo Server Is Better Than No Server: Building Blocks for Serverless Architectures
47mAdvanced Serverless Architectural Patterns on AWS
51mTaking Serverless to the Next Level
24mBuilding serverless applications with infrastructure as code
45mPatterns & practices for developing enterprise-scale serverless applications