Grid, Cluster and Virtualization -- Elastic Scaliablity (tm)
Cluster

Cacheca - Clustering Solutions

Welcome to Cacheca Homepage! Cacheca is a software company providing clustering solution to the data access layer for the most data critical and user intensive system.

Cluster was a great endeavor

Forming of computer cluster to solve computation problems has history of more than 25 years.

At the time networking was slow, memory was scared, computing power was barely sufficient and many solutions were not yet discovered. The overhead of cluster communication and operation was significant to the system. A lot of optimization must be done to a clustering system. It often required both propriety hardware and software. It was a great endeavor to get a cluster working.

Hardware is ready

Nowadays, commodity hardware has sufficient performance to handle the cluster overhead well (but not enough to handle poor cluster architecture, perhaps it never will). On the other hand, the web created a huge opportunity for interesting applications that weren't a possibility before. The industry responds to the opportunity. As advanced hardware leaps into commodity, applications that require the performance to serve millions and millions of users become a commonplace as well.

Common problems is calling for a sophisticated solution

Three tiers architecture are commonly used. In three ties architecture, there are browser client, web server (executing web application code), and Database. Occasionally, the web server tier is splitting into web tiers and business logic tiers to form four-tie architecture.

For a successful application, user scales in number itself. The web server and database server layer must scale to meet the demand of increasing browser tie (increase of users). Otherwise, the user-tie scales back itself (losing users). A lot of effort and solution has been invested and invented to scale up and scale out the database layer. It is important. A lot has also been invested in the web tie.

However, there is a bottleneck in the interface of the web server tie and the database tie: the data access layer. When the web tie (or business logic tie) consists of one machine, caching technique is used to store frequently use data in memory. When the web tie is a cluster itself, caching becomes a challenge. Sophisticated software is needed to ensure the cache contains the most up-to-date information, and it is critical to the data integrity. For this knowledge society, data is the life blood.

Cacheca's Vision

Cacheca is a provider of such a clustering solution for the data access layer. Our vision is to shift this sophisticated functionality into commodity. Cacheca adopts the dual-license model used by leading database provider (such as mysql and sleepycat). Cacheca provides both commercial license and paid supports to enterprises. Cacheca also provides its software for business and non-business use free of charge thru its open source license. Cacheca bases on a superior architecture and will meet the demand of the most data critical and user intensive system. The first version of our software is based on Java and available for J2EE platform. J2EE is the most viable web application platform as shown in all different jobs board. Alpha source is targeted to release on Q306 to the open source community. Cacheca is seeking open source developers, strategically partners and financial support. Please contact us.