Databaseless Java In-Memory Data Processing

Hibernate is the de facto standard persistence for traditional Java database apps. It was built for monolithic Java applications and also monolihic server databases. However, Hibernate is also known as complex and slow, but we got used to it. The cloud, microservices and serverless have changed Java dramatically, but persistence is still the same. Now AI is making its way into Java and AI requires uncompromising performance.

EclipseStore was built to be the next generation Java persistence for modern cloud-native systems and AI. It follows the system prevalence architectural pattern and enables databaseless in-memory data processing by using Core Java features. Different to traditional database systems that provide database-specific data structures or formats, EclipseStore supports the native Java object model for persisting objects. The engine enables seamlessly mirroring any Java object graph of any size and complexity transaction-safe into any binary object storage such as AWS S3. This leads to 3 main benefits:

1) High performance: microsecond response and query time, way faster than comparable JPA queries,

2) Significant cloud database cost savings: for example AWS S3 object storage is 70% – 96% cheaper than any database in the cloud such as AWS RDS PostgreSQL or Maria DB.

3) Simplicity: only one data model (just POJOs), any Java types are supported, no expensive mappings or data conversion, simple architecture, simple schema migration.

EclipseStore is a new Eclipse project and open source. It requires Core Java knowledges only, so getting started is easy for Java developers.