Design and Development of Albianj: A Distributed Java Framework for Large-Scale Data Access

We designed and implemented Albianj, a Java‑based distributed development framework aimed at handling massive data access and addressing the challenges of heterogeneous developer skill levels. The framework provides extensibility, customizability, and runs on simple web containers such as Tomcat or Jetty.

Albianj focuses on consistency, scalability, and ease of use, offering a single‑method solution for each requirement to avoid the complexity of multiple approaches. It adopts a "single‑dimensional" principle, reducing configuration by favoring essential settings over excessive XML configuration.

The architecture consists of eight loosely coupled sub‑modules built on a common kernel. The kernel supplies core services, IoC, thread pools, logging, and exception handling, while the IoC component uses a single interface‑based injection method to maintain consistency and simplify extension.

Data routing is a core feature, enabling transparent routing of database operations across heterogeneous storage (MySQL, Oracle, etc.) and supporting deployment modes such as read/write splitting and dual‑write. Configuration files (drouter.xml, storage.xml) define routing rules, and the FreeAlbianObjectDataRouter base class provides the implementation hooks.

Albianj includes a lightweight ORM that maps objects to relational tables without the overhead of full‑featured ORM frameworks. It integrates with the DataRouter to ensure data consistency across distributed transactions, which are managed via a job‑task model and a two‑phase commit mechanism.

The framework also implements a distributed transaction strategy that combines distributed locks with job‑task coordination, and provides a fallback "accounting" mode that records operations for later replay in unstable network conditions.

Overall, Albianj demonstrates how a custom, consistent, and extensible backend framework can improve development efficiency, maintainability, and performance for large‑scale internet applications.