Java in the Clouds: Pointless and Expensive or Productive and Must-Have?
summary
Compares concrete AWS options with Java: EC2, ECS, Fargate, EKS, RDS/Aurora, DynamoDB and Lambda, and shows a working Quarkus MicroProfile packaged as AWS Lambda from a GitHub repo; measures cost tradeoffs (one large monolith can run cheaper than many microservices, and a single lambda with 2 GB RAM often beats dozens of 128 MB lambdas), notes the control plane ≈ €100/month for Kubernetes and that the first million Lambda invocations free matters for small apps, recommends testing JVM first then trying GraalVM.
AI-written summary · about
chapters (12)
- 0:00Introduction and speaker background
- 3:00Monoliths vs microservices in Java
- 7:35Cloud expectations and realities
- 11:24Kubernetes in public clouds
- 14:17Lift and shift to the cloud
- 16:44Compute options in AWS
- 23:23Storage solutions in the cloud
- 27:13Are clouds killing microservices?
- 33:57Cost drivers in cloud computing
- 37:26Monoliths and Lambda functions
- 40:57Java in the cloud: JVM vs GraalVM
- 43:11Live demo: Quarkus Lambda example
related talks
01h 03mKeynote: Why The Future of Java in the Cloud is Serverless
56mThinking in Serverless Java (Code)
56mJava on AWS - all the details that matter
49mJava Patterns and Practices for Serverless Applications
36mServerless Java
52mContainerless, Bloatless, YAML-less, Serverless Java #slideless
53mDissecting Serverless Enterprise Java Applications #slideless
04h 47mOrange Stream