Patterns & practices for developing enterprise-scale serverless applications
summary
A real migration story: a legacy Java monolith and Oracle DB move to AWS serverless using API Gateway, Lambda, S3 and NoSQL tables plus queues/event bus for fulfillment; the experiment succeeds and grows from a few teams to 25–30 autonomous teams. Practical limits and tradeoffs are explicit: Lambda 15‑minute limit, quotas, and cost vs convenience. Recommended practices include bounded contexts, team topologies, choreography vs orchestration (in‑service, cross‑service, distributed), schemad events, observability, threat modeling and per‑team cloud accounts with CI/CD pipelines.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 2:14What serverless means
- 5:50Skills for serverless
- 10:53Enterprise scale explained
- 13:00Architecture pitfalls
- 16:32Decomposing problems
- 20:52Team topologies
- 23:00Event-driven patterns
- 27:37Choreography vs orchestration
- 32:16Development agility
- 34:47Modular apps principles
- 36:30Security in serverless
- 37:32DevOps automation
- 39:20Agility versus stability
- 40:45Key takeaways
- 42:08Q&A
related talks
53mPatterns and practices for developing enterprise-scale applications with serverless
01h 04mMigrating to Serverless - an experience report
31mWhen Bad Architectures Happen To Good People: Migrating Legacy Systems To Serverless
14mThinking Twice About Migrating to Serverless
31mAccelerate business innovation with serverless applications
49mServerless Architectural Patterns and Best Practices
01h 02mServerless in production, an experience report
42mEnterprise Serverless Adoption. An Experience Report