Sep 22, 2021 | Shaoni Ghosh
MicroStream has also made its Java persistence framework open source in conjunction with the release of MicroStream 5.0.
(Must Check: 7 Best Open Source AI Libraries)
For almost six years, MicroStream has been successfully employed in mission-critical applications. It's proven to work, reliable, and has good code quality. Anyone may now see the source code.
MicroStream was first released in April 2019. It's comparable to Java's built-in serializations, but it's a lot more capable. Unlike Java serialization, which can only persist entire object graphs, it may partly persist, load, or update object graphs on-demand. In terms of size and performance, it's also highly efficient.
MicroStream isn't a full-fledged DBMS alternative. Rather than being offered as an add-on for current database systems, the software would be better developed as dedicated server applications.
MicroStream believes that using a database management system (DBMS) is wasteful since each database has its own data structure.
As a result, data must be transformed and mapped using an extra layer such as an object relational mapper (ORM). Complexity, delay, and performance loss are all introduced by these frameworks.
The MicroStream Data-Store technology eliminates the need for these transformations, allowing everything to be stored directly in memory and simplifying the design.
They also provide data persistence connectors for databases like Postgres, MariaDB, SQL Lite, and plain-file storage (even in the cloud).
According to InfoQ, as compared to JPA, performance for a simple query using aggregation (sum) is enhanced by 10x. Data persistence connectors for Postgres, MariaDB, SQLite, and plain-file storage are also available (even in the cloud).