Terracotta Ships ''Clustered'' JVM — Simplifying Java Development; Company Announces Customer, NTT Data Intellilink, and Exclusive Distributor in Japan, Sumisho Computer Systems Corporation

Business Wire, May 16, 2006

SAN FRANCISCO -- Changing the Java landscape, Terracotta, Inc., a leader in delivering groundbreaking solutions for enterprise Java scalability, today announced it is shipping Terracotta 2.0, the industry's first production-ready "clustered" Java Virtual Machine (JVM) that simplifies Java development. The company will be debuting Terracotta 2.0 at the JavaOne Conference, Booth #408, on May 16-19 at the Moscone Center in San Francisco.

In a revolutionary move, Terracotta 2.0 clusters at the JVM level, instead of at the software application level. By doing so, Terracotta dramatically accelerates time-to-market by eliminating performance tuning from the development lifecycle, which in turn drives down total cost of ownership.

Previously, organizations have spent weeks or months writing clustering code by hand and have invested in expensive tools to make Java applications in production cluster and scale efficiently. By injecting clustering and caching into the Java runtime, Terracotta furnishes the application with linear scalability, total fault tolerance, and high availability without making any changes to the application code.

With today's announcement, NTT Data Intellilink, a subsidiary of NTT Data, the largest systems integrator in Japan, is among the first customers of Terracotta technology. Also announced today is Sumisho Computer Systems Corporation (SCS), the exclusive distributor of Terracotta products in Japan.

"Organizations, such as financial services and telecom firms, are increasingly engaging in business critical J2EE projects. As their business grows, users need low cost ways to scale their applications up quickly," notes Massimo Pezzini, Vice President Distinguished Analyst, at Gartner, Inc. "To support the scalability, performance and availability demanded by bet-the-business, multi-node, enterprise Java applications, developers and operators are looking for solutions seamlessly combining high performance clustering, grid architectures and caching services while requesting minimal intrusion and re-architecting of the application code."

Adds Ari Zilka, founder and CEO of Terracotta, "The reality is that clustering is a challenge everyone faces, from Fortune 500 companies to organizations running a couple of Tomcat servers. Terracotta makes low-cost, drop-in clustering available to anyone, regardless of industry, regardless of organization size."

Terracotta Defines the "Clustered" JVM

Terracotta defines a "clustered" JVM as presenting an unlimited number of hardware servers (CPU's, RAM, network, and hard disk included) as one JVM to a business application, thus providing an unlimited amount of memory and compute power to meet business needs. As a configuration-driven clustered JVM, Terracotta 2.0 plugs into off-the-shelf JVMs, letting the user target exactly what resources to cluster, such as objects and locks.

Terracotta 2.0 provides significant benefits to:

--Power users that want to cluster by hand, but without any special clustering code, such as large financial services and telecommunications firms using a grid computing framework.

--Framework and open source users who have adopted well-known frameworks, such as J2SE, Spring, or Hibernate and just want to write applications without worrying about the framework's clustering capabilities.

Terracotta 2.0 includes Terracotta Distributed Shared Objects (DSO) and Terracotta JDBC -- the underpinnings of the clustered JVM technology, as well as the introduction of Terracotta Sessions.

Terracotta DSO is a runtime solution that shares data across JVMs without APIs, custom code, databases, or message queues. Terracotta DSO clusters the JVM so applications are deployed across a large number of servers in a fault-tolerant architecture. Terracotta 2.0 introduces an Eclipse plug-in for quickly configuring objects and services to be clustered, as well as an enhanced management console for monitoring and tuning.

Terracotta JDBC is a runtime solution for clustering JVMs to keep an in-memory copy of data that is automatically synchronized with the database, significantly reducing calls to the database. Terracotta JDBC learns schemas and monitors database changes as a completely drop-in solution. Terracotta 2.0 introduces constant-time query cache, row-level cache invalidation to further improve database utilization, pattern-based opt-in/opt-out of queries for caching, and an improved console with cache tuning capabilities.

Terracotta Sessions is a simple, drop-in product that dramatically improves HTTP session clustering performance and availability. Terracotta Sessions is available for free and can be deployed in 30 minutes. Terracotta Sessions is currently available for Apache Tomcat and BEA WebLogic Server.

Free downloads are available from the Terracotta website:

--Terracotta Sessions for Tomcat and Terracotta Sessions for WebLogic Server at http://www.terracottatech.com/downloads.jsp.> Related News

May 16, 2006 -- Terracotta Begins Free Downloads for Java Developers (http://www.terracottatech.com/press_5_16_06_download.shtml)