Business architecture

A three-tiered architecture includes one more node between the client and the database server—the middle tier. In a three-tiered architecture, business logic is offloaded from the client and the database server nodes to the middle tier. In fact, you can choose to further distribute the business logic among more than one middle tier node and still call it a three-tiered (or K-tiered) architecture because the idea is similar. Note that the tiers do not have to be physically separated. You can have both the middle tier server and the database server collocated on the same physical computer but running in different processes with separate memory spaces. Modem hardware architecture can partition a single hardware box into multiple virtually separate computers or domains. Typically, a three-tiered architecture supports a Web-based thin client although it can also work with a rich client. 

Quid, M. A., Designing a Reengineering Proof Process Architecture, Business Process Management Journal, Vol. 3, No. 3, 1997, pp. 232-247. 

The most commercially successful of these systems has been the Convex series of computers. Ironically, these are traditional vector machines, with one to four processors and shared memory. Their Craylike characteristics were always a strong selling point. Interestingly, SCS, which marketed a minisupercomputer that was fully binary compatible with Cray, went out of business. Marketing appears to have played as much a role here as the inherent merits of the underlying architecture. 

IBM Program Product MVS/Extended Architecture Operations International Business Machines, 1984. 

We have developed an approach termed ARBITER (Architecture for Reliable Business Improvement and Technology Evaluation in Research) [25], with associated simulation and visualization capability, that combines... 

A component, on the larger scale, often supports a particular business role played by an individual or department with responsibility for a particular function. Businesses talk increasingly of open federated architectures, in which the structure of distributed components mirrors the organizational structure of the business. When reorganizing a business, we need to be able to rewire software in the same way. 

Application architecture The major components, how they partition the business logic, and how they interact, defined primarily with collaborations and types. 

Make an early architectural decision about how much you will tradeoff performance, seamlessness, reuse, code flexibility, and so on. If your clients shout for little functional enhancements every day (something that is typical for in-house financial trading software), optimize the underlying communications and infrastructure but leave the business model pristine. But if your software will be embedded in a million car engines for 10 years, optimize for performance. 

The purpose of packages is to make explicit the dependencies between different areas of the development work. A variety of patterns can be apphed to help separate concerns, including vertical separation of different views, horizontal separation of different architectural and business levels, and code separations of interfaces from implementations. 

When you build a design from components, you don t need to know how they are represented as objects or as instances of a classes or know how the connectors between components work.2 In federated systems, just as in 00 programs, each component is a collection of software it is chosen for the support it provides of the corresponding business function and uses local data representations best suited to the software. Just as in OO programs, objects must access the information held by other objects, so in a component architecture, components intercommunicate through well-defined interfaces so as to preserve mutual encapsulation. 

Server components typically implement significant business functions and mn on a server. In a multitier architecture, most business logic in an application runs on dedicated servers... 

In practice, it is useful to distinguish the application architecture—how the business logic is split across components and how they interact—from the technical architecture all the infrastructure and other domain-independent pieces that support that collaboration. The four-tier Web-enabled architecture presents a typical case for making this distinction. 

In our view, architecture is not only about Gothic-scale structures but is also about all structures and relationships used down to the level of code.1 The decision to use a four-tier structure, with a thin client, a Web server, a business application server, and a database, is architectural. But, in the extreme, we consider a consistent use of getXQ and setX(x) methods also to be part of the (detailed) architecture. This view leads to a somewhat less formal definition of architecture. 

Use a three-tier client-server architecture. All business logic must be in the middle tier, presentation and dialog on the client, and data services on the server. In this way you can scale the application server processing independently of persistent storage. 

The 1.0 specs of Enterprise JavaBeans are a good example of standard architectural patterns and how they can be used to define a simple and consistent architecture even for large-scale business systems. 

Unfortunately, until component-based development becomes the norm and until project managers understand the economics of buying the kinds of domain-independent components described earlier, we will still be building many of them. If most of the elements of the technical architecture are already implemented, then estimating the development time for business functions becomes much less like a black art. 

A typical Web-enabled business system might adopt an architecture with four physical tiers, as shown informally in Figure 12.5. The four tiers are as follows. 

The extent to which the code reflects the business model is related to the architecture. If the code must be changed every few days to keep competitive with others (for example, financial dealing systems), the requirements of users should translate directly to code changes, which must closely reflect a suitably flexible business model. If high performance is needed and changes are rare (as in an... 

Documentation is structured around specifications and implementation and their refinement and import relationships. A user manual—a description of how a user accomplishes tasks by using the system(s)—is a particular form of documentation associated with a refinement how the abstract business model and actions are realized by more-detailed actions performed by the users and systems. Test specifications are also associated with refinement relationships. The rules for system architecture are documented in a package that specifies the patterns and frameworks that will be used in other packages that import it. 

Using refinement, you can model objects and interactions at all levels of granularity. This arrangement provides a fractal view of architecture, from the business roles and processes to large-grained interacting architectural components including a system to individual interfaces and classes. Any refinement has associated architectural decisions. 

The application architecture A package structure and collaborations. This implements the business logic itself as a collection of collaborating components, with the original specification types now split across different components. The components can range from custom-built to common off-the shelf components, such as spreadsheets, calendars, and contact managers, to purchased domain-specific components such as factory-floor schedulers. This application architecture lives atop, and uses, the technical architecture. 

The technical architecture Apackage structure (for static dependencies) and collaborations (across technology components, such as UI, business object servers, and databases). These cover all domain-independent parts of the system hardware and software platforms infrastructure components such as middleware and databases utilities for logging, exceptions, start-up, and shutdown design standards and tools and the choice of... 

The business case This provides initial requirements, defining the business problem or opportunity that this project addresses. It typically includes a high-level list of numbered functional and nonfunctional requirements2 (lb), the business reasons and risks for the project (la), the scope of the project in terms of things definitely included or excluded, linked clearly to a business model in terms of business objectives, actions or use cases, and user roles that must be supported (la), known requirements on the architecture, design, and implementation (lc), and constraints on project budget and schedules (Id). 

In this case study we found it useful to distinguish six descriptions a different project might make a somewhat different separation. The primary separation of domain or business (the outside), system specification (the boundary), and internal architectural and detailed design (the inside) will always hold. The following two features are common to each level. 

Pattern 16.7, Implement Technical Architecture Define major technical components of your design as an architectural collaboration. These components might be GUI, business logic, persistence (database or file system), communications, and other middleware elements. 

The GUI is a separate layer. In contrast to an older client-server (two-layer) architecture, the user interface should deal only with presenting business objects to users and translating user typing, mouse clicks, and so on to business object commands. Business rules should be embodied within the business objects. 

Figure 16.16 Separating store systems and business systems in the architecture. 

This project n/as developed specifically for the new Moscow International Business Center urban site reacting to the architectural context and dimensions. Two stylized heart chambers, whose partition moves hydraulically, pulsate as a result of the continual compression and expansion of stainless steel fins. 

Architecture. Many common practices negatively affect SCADA security. For example, while it is convenient to use SCADA capabilities for other purposes such as fire and security systems, these practices create single points of failure. Also, the connection of SCADA networks to other automation systems and business networks introduces multiple entry points for potential adversaries. 

Garud, R., Kumaraswamy, A., Malhotra, M. 2003. Infosys architecture of a scalable corporation. Stern School of Business Case, New York University. 

---