High Performance Java Cloud Native Apps
summary
EclipseStore (formerly MicroStream) replaces databases by keeping Java object graphs in memory and persisting only micro‑snapshots of deltas to blob storage, with lazy loading, transaction‑safe store calls and Java Streams queries. A bookstore demo shows queries up to 1000x faster than JPA and claims ~98% cloud DB cost savings (example: a replicated 6‑node Postgres cluster ~$50k/yr vs object storage ~$70/yr). The stack uses the Eclipse serializer, Java 8+, optional uses AWS S3 or plain files, runs as a JAR on Kubernetes and is open source.
AI-written summary · about
chapters (14)
- 0:00Introduction and speaker intro
- 1:56Cloud storage costs pain point
- 5:31Evolution of cloud and microservices
- 8:44Java challenges with microservices
- 14:29Solutions with micro runtimes
- 17:48GraalVM native images trade-offs
- 19:36OpenJDK Project CRaC for JVM startup
- 23:42Traditional Java persistence problems
- 29:04High performance with pure Java
- 34:08System prevalence and Eclipse Store
- 38:49Eclipse Store performance demo
- 41:31Distributed apps with Eclipse Store
- 45:30Eclipse Store code simplicity
- 50:07Conclusion and benefits summary
related talks
56mEclipseStore - Databaseless Java Cloud Native Persistence for Microservices
53mDatabaseless Data Processing - High-Performance for Cloud-Native Apps and AI
54mEclipse MicroStream - Ultra-Fast Java Cloud-Native Persistence for Microservice
01h 00mEclipse MicroStream - Ultra-fast Java cloud-native persistence for microservices
43mThe Java Cloud-Native Stack for Microservices and Serverless Architecture
52mGet Started Coding with EclipseStore...
55mDatabaseless High-Performance Persistence with EclipseStore...
43mEclipseStore - Ultra-Fast Java Cloud-Native Persistence for Microservices and Serverless