Latest News

08 May 2009 Alpha version of HXSQL 1.9 (feature compatible with HSQLDB 1.9) is undergoing tests. Release version will be available at the same time as HSQLDB 1.9.

 

The 1.8.1.x series, which will be fully compatible with the latest HSQLDB 1.8.0.x, will feature additional performance improvements over HXSQL 1.8.0 and is scheduled for release in June.

 

Current release versions of HyperXtremeSQL feature all the latest enhancements to HSQLDB plus dedicated support for multi-gigabyte data sets and fast database operations in a very small memory footprint.

Currently available versions include:
Gereral-purpose version 1.8.0.10 for use with JDK 1.4 or later.

Sepecial PDA version 1.8.0.10 for use with JDK 1.1 - JDK 1.3.

Special version 1.8.0.10 for the CDC embedded platforms.

Introduction

HyperXtremeSQL™ is a super-fast SQL database engine written in Java. It support SQL and JDBC in J2SE and J2EE contexts.

HyperXtremeSQL is fully compatible with the open-source HSQLDB engine at the application level and can replace the HSQLDB jar in an existing application with no change to the JDBC Java code or SQL queries. It can be used as an embedded java database engine or as a database server. HyperXtremeSQL is built around a new indexing and persistence core that replaces HSQLDB's core classes. The HyperXtreme core breaks new ground as it lowers the code footprint, reduces memory use and at the same time significantly speeds up query performance in all areas of SQL, all modes of database operation and all sizes of database.

HyperXtremeSQL is lightweight, featuring full embedded functionality at around 350K jar size and full client-server functionality at around 400K.

Since its introduction in late 2004, HyperXtremeSQL has been in use on thousands of desktop machines and tens of thousands of embedded platforms both by end users who upgraded from HSQLDB and by ISV's as part of their application programs.

^ TOP

Speed

HyperXtremeSQL raises the SQL performance bar to such levels that allow existing applications to scale up beyond their original limits. In many database usage scenarios, SELECT speed improvements between 600% and 1100% over HSQLDB are achieved within the same memory footprint.

Rows of data can be selected at rates of up to hundreds of thousand rows per second and inserted at tens of thousand rows per second. By harnessing such phenomenal performance, HXSQL opens new horizons for all Java applications and allows sophisticated features and much higher levels of functionality to be achieved on all platforms.

• Server High-performance server applications are no longer hampered by slow database layers. No need to implement complicated proprietary data access and persistence schemes. Standard SQL and JDBC can be used in all areas.
• Desktop Data-intensive workstation applications can benefit from the power of relational queries and OLAP. With the use of SQL, development efforts can be channelled to the implementation of domain-specific features instead of data access and manipulation.
• Embedded Embedded devices with limited speed and memory will benefit from the speed increase and smaller memory footprint.

The SQL / JDBC interface is highly optimised and adds only a marginal overhead to data storage and retrieval. Applications that need fast indexed access to large amounts of data, whether in memory or on disk, can use HXSQL instead of bare indexed access solutions. The fully-featured SQL query engine offers additional query possibilities to simplify reporting and management.

^ TOP

Scalability

HyperXtremeSQL can scale up or down to suit the platform and data set. By default, it can handle millions of rows of data with up to 16GB storage (storage can be raised to 128GB and beyond if required for the specific usage). It has new modes of data backup and checkpoint for seamless operation with huge data sets. Special versions scale down for use in embedded platforms (PDA and embedded applications), offering maximum speed in a small memory footprint.

^ TOP

Background

HyperXtremeSQL is a separate product from HSQLDB. It has been developed by Fred Toussi, lead developer and maintainer of HSQLDB. HyperXtremeSQL will continue to be developed further alongside HSQLDB and will share the code outside the database core. Sales of HXSQL licenses and support services will contribute to continued development of both products. The first release of HyperXtremeSQL was launched in November 2004.

HSQLDB is the leading open-source SQL database engine written in Java and features extensive support for SQL and JDBC. The engine enjoys widespread trust and support among Java developers and has been included with many software products and applications, both open-source and commercial. It is a proven solution and the fastest SQL engine compared to open-source alternatives when accessed from Java for many common types of SQL queries. It has been developed, debugged and refined continuously in the last 7 years with over 60 alpha and beta releases and 7 final releases.