Like most software terms, the term virtualization is becoming hopelessly overloaded.  As pointed out in a recent article, a definition of the different types of virtualization can be hard to come by.  When people see the Appistry Enterprise Application Fabric (EAF) solution one of the first things they think of is application virtualization.  In this post I will try to resolve the confusion by detailing some of the key EAF features and how they relate to virtualization.  We will see that the EAF properties that promote scale, reliability, and maintenance of your fabric are better known as service virtualization.

An application virtualization definition

This is a definition of application virtualization as provided by the referenced link:

Application virtualization provides the possibility to fool an application into believing that some or all of the system resources it accesses and tampers with are actually the real deal, while they are in fact simulated by the application virtualization management software.

By wrapping a single application within an application container (such as MAV or XenApp) a desktop application can be moved easily to any machine.  A system administrator no longer has concerns of installing and configuring a single application to run on a large number of machines.  The application container and application management software will take care those deployment problems.

This is a great tool for a system administrator to have at their disposal, but it addresses the problem of desktop application installation and remote application invocation.  These are not the problems that Appistry's fabric solves.

Appistry EAF and service virtualization

EAF provides scale, reliability, and maintenance features by recognizing that the operations that comprise your application can really be thought of as services.  The fabric is able to distribute these services across the machines in the fabric and virtualizes them to look like one system.

One of the base principles of EAF is to make a large number of computational resources appear as a single entity to an application developer or administrator.  The fabric learns about your application, through the declarative statements provided by the developer and/or administrator.  Let's illustrate this service virtualization with several of the fabric's features:

• Fabric application model: A typical EAF application consists of two main components; the binaries that embody the business logic of your application and the declarative model that instructs the fabric how your application ties together.  Once the fabric understands the construction of your application, each worker becomes a self-sufficient provider of those application services.  Whenever your request results in a need for a task, component, or process flow to be executed, a worker in the fabric can locate another worker to provide that service.
• Auto-discovery, self-organization, and scalability: EAF automatically discovers and organizes the machines within a fabric.  When a new worker is detected, the worker will be updated with the latest system, application, and configuration files. The worker will also be added within the fabric's communication structure to make itself available to run the components of your application.  Your application's services become seamless extended to other machine in the fabric.
• Fabric management: The fabric's operation layer is responsible for management activities. Whether the operation is a simple worker status query or application update an administrator simply pushes the request to the fabric and lets the management layer take care of the rest.  All the machines of the fabric are virtualized as a single resource.

In conclusion

EAF takes advantage of service virtualization to weave together large numbers of computers into a single computational fabric.  This virtualization greatly facilitates the job of the application developer and fabric administrator.  By depending on the fabric, application developers get to focus on the business logic of their application.  By depending on the fabric, administrators know that they have a manageable infrastructure which is able to painlessly scale to their needs.