Evolve your web application for scale
summary
Starts from a single EC2 + Elastic IP monolith that suffers a two‑hour outage during a MySQL upgrade and evolves into a multi‑AZ, multi‑Region web app using CloudFormation, ALB, Aurora (MySQL/Aurora Serverless), RDS Proxy, ElastiCache for Redis (sub‑millisecond), S3 + CloudFront (450+ edge locations), CloudWatch Synthetics, DevOps Guru, Route 53 ARC, Auto Scaling and AWS Fault Injection Simulator. Outcomes: reduced downtime, session reads moved to Redis, connection storms solved by RDS Proxy, storage offloaded to S3, autoscaling cuts cost and carbon (Aurora ~10% premium). Aurora replica with synchronous cross‑AZ failover CloudFront to S3 and ALB
AI-written summary · about
chapters (13)
- 0:00Introduction to evolving architectures
- 1:16Starting with a monolith example
- 2:23Improving reliability with automation
- 3:52Moving databases to RDS Aurora
- 5:20Adding high availability with ALB
- 8:45Addressing performance issues
- 11:40Using DevOps Guru for diagnostics
- 14:59Optimizing for sustainability
- 18:42Testing resiliency with chaos engineering
- 19:47Securing credentials with Cognito
- 21:30Expanding to multi-Region deployment
- 26:37Key takeaways from the evolution
- 28:46Closing and resources
related talks
40mCloud Databases: Pros and Cons
44mIntroduction to application scaling
54mAWS Certified Solution Architect an Overview - Part 1
41mSeven patterns to increase application availability
47mFrom Legacy to High-Load: Evolution of a Web Application
32mBeyond Optimization: Resilient Cloud Microservices at Scale
30mHow We Engineered Value in Times of COVID?
35mAWS Lambda performance tuning