Rosella DBMS for Data Mining and Data Warehousing

Rosella is a large main-memory optimized high performance Java cross-platform disk-based relational DBMS, featuring real-time replication, advanced disaster recovery, MM-DBMS level speed, statistics, data mining, hotspot search, data warehousing, embedded operation (from Java programs, Servlets, JSP, Tomcat, JBoss, WebLogic, WebSphere, ...), and supporting JDBC, ODBC, SQL / CLI, J2ME connections ...;

Statistics, Data Mining and Data Warehousing

Warp-speed performance of Rosella backed by ultra fast sort engine makes Rosella ideal DBMS for data mining and data warehousing. In this end, Rosella support various statistical and data mining functions.

Warp-speed Performance

Rosella is optimized for large main-memory. Its single buffer architecture allows not only caching a large number of data pages, up to 512GB, but also handles large sorting data that can take advantage of the ultra fast radix sort built into Rosella. For example, it allows Rosella to sort a million records within a few dozen seconds!

100% Java and 100% Portability

Rosella is written in Java 100%. It can run on various platforms: Windows, Mac OS X, Solaris, Linux, ... Furthermore databases created on a platform can be ported onto other platforms without costly conversions! Simple file copy will do the porting!

Embedded Database Systems

Being 100% written in Java, Rosella can be integrated into user's J2EE web-servers or Java business applications. Embedding Rosella DBMS is as simple as adding an archive file and calling start/stop APIs. Note that embedding greatly improves performance of your system by virtually eliminating communication cost.

Crossplatform Realtime Replication

Rosella provides various data recovery mechanisms. For example, dual writes, dual drives, safe writes, and replication. By far the most powerful one is realtime multiple replication over different platforms. This not only prevents major data loss from disasters and computer virus attacks. But also it removes the need to make daily or periodic backups!

High-level Concurrency and Transaction Support

Rosella employs advanced locking and concurrency control mechanisms. Especially high degree concurrent retrieval and updates with indices will provide high level concurrency for demanding applications.

Industry Standards Supported

Rosella supports the following industry standards to make it easy to migrate from/to Rosella. This will assure compatibility with other relational DBMSs;

• SQL-99 for standard programming.
• JDBC for Java applications.
• TCP/IP for network connections.

Enterprise Business Intelligence and Data Warehousing

Most database systems support only a very limited set of statistical functions. On the other hand, statistical packages and OLAP tools are very limited in managing data, compared to database systems. Rosella and associated systems overcome these limitation combining both database and statistical analysis together;

(Example) Realtime Risk Management Dashboards;

(Example) Data visualization and Charts;