About Us
Products
Industries
Services
Partners
Media Room
Support & Education
Testimonials
Request Info
Become a VAR
Support
Products
HMS
HRMS
FMS
Caché
TrakCare
Labtrak
Case Studies Downloads Careers Search Site Map Reach Us Print this page HOME
Go Home
        Download Cache

Caché Executive Summary

Caché Architecture

The hallmarks of Caché, the post-relational database, are high performance, massive scalability, rapid application development, and cost-effectiveness. These values are reflected in Caché's basic architecture.

Caché stores data in an extremely efficient multidimensional form, thus ensuring lightning-fast performance, even under heavy loads running on significantly less hardware than other database systems. Plus, Caché makes its data accessible through a wide variety of technologies, which promotes both openness and rapid application development because developers can work with familiar, readily-available tools.

The Multidimensional Data Engine

Unlike relational databases, which force data in two-dimensional tables, Caché stores data in multidimensional arrays. In addition to enabling realistic data modeling, multidimensional arrays are much faster to access because they eliminate the processing overhead associated with "table-hopping" and "joins" that typify relational technology. Another performance-enhancing feature is Caché's unique Distributed Cache Protocol, which dramatically reduces network traffic in distributed systems. In customer tests, Caché has performed up to 20 times faster than relational databases.

Although data is stored in multidimensional form, Caché gives developers the freedom to model their data any way they choose: as objects, as tables, or as multidimensional arrays. Caché comes with an easy-to-use graphic interface for creating Caché Objects. It can also accept input from Rational Rose (an object modeling tool) and DDL files (the standard for defining relational tables).

By virtue of Caché's Unified Data Architecture, all data is automatically accessible as both objects and tables. There is never a need to "map" from one form to the other, and no processing overhead required to convert between forms. The Unified Data Architecture increases both productivity and performance.

Caché also allows choices when it comes to database and business logic scripting. Caché ObjectScript supports all data access methods: objects, SQL, multidimensional, and even embedded HTML. Caché Basic is similar to Visual Basic, with a few modifications to take advantage of unique Caché capabilities.

Web Access

In keeping with InterSystems' core values, Caché's Web connectivity is geared toward providing high-performance and massive scalability, coupled with a super-fast application development platform. In Caché's unique Web architecture, Caché Server Pages execute on the data server, close to the data they need to access. Not only does this approach boost performance, it greatly enhances scalability by taking much of the processing load off the Web server, leaving it free to handle more browser requests.

Caché applies the rapid development power of object technology to the creation of Caché Server Pages. Every Caché Server Page is itself an object, and can inherit session management behavior (of various levels of security) from system objects provided by InterSystems. This frees application developers from much of the tedious system-level coding needed to maintain "state" during user sessions. Object inheritance is also a quick way to ensure a consistent "look" across all pages of an application.

Additionally, Caché simplifies Web development by allowing Web designers and application developers to work in parallel to achieve the final result. Using familiar and readily-available Web authoring tools, Web designers add functionality to pages by incorporating Caché Application Tags (CATs) the same way they would any standard HTML tag. CATs for some standard functions come with Caché, or they can be custom-built. Application developers can write CATs that perform useful functions, regardless of the design of the page that contains them. As a result, Web applications can be developed more rapidly and efficiently to enable the short time-to-market that's essential on the Web.

Object Access

These days, virtually all new application development is done using object modeling techniques. Modeling data as objects allows developers to think about data in a natural, intuitive way. And because objects are modular, with well-defined interfaces, they are reusable, and can be shared between applications, resulting in significant productivity gains.

Caché supports a full range of object modeling techniques, including multiple inheritance, encapsulation, polymorphism, references, collections, relationships, and BLOBs. Caché Objects can be created with the Caché Object Architect (an easy-to-use graphical interface) or through Caché's bi-directional link to Rational Rose (a popular object modeling tool). Unlike some "object-relational" database systems, Caché allows data schema evolution, so object definitions can be altered to fit the changing needs of your applications. And, thanks to Caché's Unified Data Architecture, all Caché Objects are automatically ODBC-compatible.

Caché Objects are also compatible with a wide range of object-oriented tools and technologies. They can be used by Java and C++ developers, and by tools (such as Visual Basic and Delphi) that use the COM interface. Caché also comes with a bi-directional CORBA interface.

SQL Access

In their heyday, relational databases were ubiquitous, and even today they represent a majority of the databases still in use. Many software applications, particularly those for data reporting and analysis, use SQL as their query language, and require an ODBC- or JDBC-compliant database at the other end. Through it's SQL data access, Caché is available to all these applications. In addition, the Caché SQL Gateway allows Caché applications to access data stored in relational databases - very useful when there is a need to integrate data from a variety of sources.

Some developers may wish to migrate applications from a relational database to Caché in order to take advantage of Caché's higher performance and advanced object technology. Caché can create data structures from relational table definitions contained in DDL files. By virtue of Caché's Unified Data Architecture, every table definition becomes a simple object that can be used as is, or as building blocks for more complex structures. Then, using the SQL Gateway, data can be transferred from the old relational database into Caché.


Multidimensional Access

Multidimensional data access provides compatibility with other InterSystems products, all of which use the same kind of multidimensional data structures as Caché. MVON is a MultiValue Basic implementation that, through Caché's multidimensional access, can point to Caché as its data repository.

 

Learn More
Caché Features & Benefits
Caché Tech. guide
Components of Caché
The Powerful Fusion of Objects & SQL in a Multidimensional Engine
Why Caché is the post-relational database for Web apps
Caché Executive Summary
Cache Current Releases