Databaseless High-Performance Persistence with EclipseStore...
summary
EclipseStore (formerly MicroStream) replaces JPA/Hibernate by keeping Java object graphs in-VM, snapshotting deltas as appended binaries to plain files or AWS S3 and rebuilding lazily on demand; demo shows typical queries 10–100x and sometimes 1000x faster than JPA with microsecond response time, and claims 98% cost savings by eliminating DB clusters and caches. It uses the Eclipse serializer, a single store(owner) call, replication with eventual consistency for clusters, legacy type mapping for schema changes, a storage garbage collector, and a closed-source managed cluster/SaaS offering.
AI-written summary · about
chapters (10)
- 0:00Introduction and speaker background
- 3:05EclipseStore project overview
- 13:19Problems with traditional databases
- 20:10System prevalence and EclipseStore
- 25:10Performance demo introduction
- 28:47Distributed applications with EclipseStore
- 31:18Paradigm shift in persistence
- 35:14Core functionality and usage
- 43:27Future developments and roadmap
- 45:57Q&A and project support
related talks
22mEclipseStore – The new, ultra-fast Java persistence for the cloud
43mEclipseStore - Ultra-Fast Java Cloud-Native Persistence for Microservices and Serverless
56mEclipseStore - Databaseless Java Cloud Native Persistence for Microservices
53mDatabaseless Data Processing - High-Performance for Cloud-Native Apps and AI
01h 00mEclipse MicroStream - Ultra-fast Java cloud-native persistence for microservices
53mHigh Performance Java Cloud Native Apps
54mEclipse MicroStream - Ultra-Fast Java Cloud-Native Persistence for Microservice
44mCaching Unleashed: Revolutionizing Java Performance
