E-Commerce Technologies

A growing variety of technologies

A growing variety of technologies have been developed over the past few decades with the intention of integrating multiple distributed applications in a network. The majority of networks in a normal business environment are diverse and comprise several groups of machines and operating systems. Components in such dispersed enterprise systems must communicate with one another remotely via a variety of methods. However, communication has become difficult due to a lack of interoperability and divergent operating standards. Integration activities are required to foster collaboration, enhance efficiency, and gain a competitive edge in the market as distributed systems continue to increase in complexity and size as a firm expands. Web services have been developed to promote the integration of disparate platforms and overcome existing challenges. Consequently, the following report will highlight on how Fluency English Services Ltd. (FESL) can leverage from web services, in particular, Service Oriented Architecture (SOA), to enhance performing lookup operations on their current systems.

Adopting a Service Oriented Architecture (SOA) Based Approach

Adopting a Service Oriented Architecture (SOA) based approach represents a situation where an enterprise utilizes web services to integrate heterogeneous and distinct systems in a network. It represents an architectural concept that supports the transformation of a business into a set of interrelated services. Accordingly, it also offers guidelines on how services are discovered, described and used. Versteeg (2008) argues a service represents either a logical representation of a repeatable business activity with a specified outcome that is also self-contained. Furthermore, He and Da Xu (2014) reiterate that these services can be interrelated through interfaces, thus making them independent of hardware and operating systems of the computing platform. This allows different services to interact with others in a universal and uniform manner, and thus making them cross-platform. Another great benefit that arises from such services is because they can be reconstructed and reused to develop entirely new applications (Dumitrache, Dumitra & Baciu, 2010). Therefore, service-oriented integration can be described as an Enterprise Application Integration (EAI) standard based connection that offers flexible sets of routing, transformation and monitoring capabilities. These services coalesce to accomplish a specific task, where businesses can adapt quickly to the rapidly changing requirements at reduced costs.

Analysis of Tiered Architectures

In most cases, architectures of traditional or legacy distributed systems follow a single- or two-tier architecture, where a tier is a logical layer of a system (Versteeg, 2008). Separating applications into multiple tiers is underlined as a good practice, based on its ability to reduce coupling, and making it simpler to change or replace parts of a system. Early generation enterprise applications were built on centralized mainframes, using a monolithic or single tier, making them time-consuming and hard to maintain. Consequently, this led to the emergence of two-tier client-side architectures, where the client is responsible for presenting the application to the user, while storage and data management were handled by the server. Client applications in two-tier architectures are called "fat client" because they handle data processing locally. Nevertheless, two-tier architectures were costly and limited in functionality, which led to the introduction of multi-tier client-server architectures. Coupled with the introduction of the internet, Dumitrache, Dumitra, and Baciu (2010) insist that it was then that the multi-tiered architecture developed as the de facto computing platforms till the mid-2000s.

Differences between SOA and Tiered-Based Approaches

From a physical perspective, both service-oriented architectures and legacy internet architectures are similar. However, the difference between the two architectures is the separation of concerns (Versteeg, 2008; Ahalt & Krishnamurthy, 2002). In tiered architecture, most distributed internet systems place all their application logic on the server-side and thus centralizing all the operations. Therefore, the major differences between SOA and Multi-tier architectures are based on logic design: Partitioning of the business logic, presentation layer, communication layer and methods of data integration (He & Da Xu, 2014).

Business Logic Integration

Often, the traditional tiered architecture applications do not decompose business logic into independent pieces, and integration at this layer normally consists of operating systems and network programming. Therefore, the resulting system becomes difficult to upgrade, maintain and configure (Versteeg, 2008). In the case of SOA, application integration at the business logic is enhanced through the use of middleware technologies to integrate distributed enterprise applications. Middleware technologies are developed to hide-lower level machine primitives, which makes the design of distribution systems much faster and easier. Furthermore, Ahalt and Krishnamurthy (2002) insist that abstractions offered by middleware allow the isolation of applications from the constantly changing computing components.

Presentation Layer Integration

In SOA, integration at the presentation layer normally focuses on the integration of the user interface (UI), with the goal of presenting a dynamic and integrated view for users. For traditional tier-based applications, proprietary and platform-dependent interface technologies are incorporated into the overall logic, making them suitable for specific clients (He & Da Xu, 2014). Platform independent technologies like CSS and XHTML are adopted to enhance portability, thus making it possible to run two user interfaces without duplicating the existing business and application logic. Though Dumitrache, Dumitra, and Baciu (2010) acknowledge the progress made on integrating the presentation layer, they, however, recommend undertaking further research since standardization remains a problem.

Communication Layer Integration

Interfacing of systems in traditional tier-based approaches is normally dependent on accessing an application's database through running SQL queries on it. This leads to the creation of tightly coupled dependencies that make such systems resistant to change. In SOA, integrating distributed applications ensures that these systems can maintain constant communication and exchange information when necessary. Jestratjew and Kwiecien (2013) also underline that transport of information between layers is done through HTTP and IIOP protocols, which are more secure since they pass through a firewall.

Data Integration

Data Integration involves the movement of data between heterogeneous data systems residing on different machines, operating systems, and databases. In SOA, this involves a lot of data conversion in the source and target schema, where a mapping provides queries for such operations. Versteeg (2008), however, points that maintenance and understanding of such data systems is complicated. In traditional systems, there are no explicit methods to integrate data between heterogeneous systems except through databases.

Modifications FESL would need to undergo to Integrate with the Web Services Model

Through adopting SOA-based Enterprise Service Bus (ESB), the company would not require modifying any existing applications to integrate with the web service model. This is because adopting ESB, with SOAP, an open protocol, would shield the company from different proprietary protocols for their heterogeneous systems. Consistent with He and Da Xu (2014) views, this would enhance the interoperability and flow of data between different systems. In the case of FESL's current IT setup, it is evident that its databases use distinct formats and exists in different operating systems. Since most of its branches span over 100 regions, coordination between different staff becomes a challenge due to lack of standardization. However, by using SOA-based ESBs middleware technology, users can receive SOAP requests and XML documents; allowing them to manipulate them based on explicit and standardized rules (Jestratjew & Kwiecien, 2013). Irrespective of the incoming requests, non-XML documents can be converted to the standard XML format. Consequently, the transformation component makes it easier for the company to standardize on the local currencies of its employees, using pre-determined exchange rates. Since FESL does not require modifying existing applications to upgrade its database systems, reusing existing systems drives integration and help business respond to market more cost-effectively (Ahalt & Krishnamurthy, 2002).

Disadvantages associated with Adopting SOA

Security. As a result of the increasing adoption of Web Services in organizations to integrate various systems, security challenges are becoming prominent. Since SOA normally uses open standards, services may thus be more open to applications when compared to a monolithic application. Though White et al. (2012) acknowledges that WS-Security deals with such challenges, he recommends that more research must be undertaken.

Reliability of Service-oriented integration. Simanta et al. (2012) pointed out that reliability is also another challenge, where the chance of such applications collapsing is high. Current load balancing, transaction strategies, and fault-tolerance capabilities are below the threshold, which reduces their robustness and maintainability.

Quality and Integration Difficulties. Based on the need for web services to provide accurate and consistent information, selection web services from diverse services is a challenging task. In particular, Khadka (2013) highlights that most third-party web services are unreliable as most of them are badly coded. Integration of applications is also complicated as a result of lack of guiding standards, where maintenance and compatibility with existing applications are difficult (Simanta et al., 2009).

Selecting the correct strategy. Though there exist some strategies to transform legacy systems, Khadka (2013) research found out that the time and costs of migrating SOA are difficult to estimate. This may lead to reuse difficulties and staff retraining due to lack of requirement understandings.

Benefits of SOA

It can be observed from the case study that the main benefits associated with the adoption of SOA include allowing interoperability of applications and services by allowing data exchange between distributed applications. Therefore, users in one of FESL's branches can be able to access customers' records at the main branch in New Zealand without having to upgrade their applications. Concisely, Business Process Management (BPM) is a strategy that is used to manage the performance of a business through optimizing business processes. Therefore, BPM is used in an organization to learn, make decisions and develop business process software (Aguilar-Saven, 2004). BPM is thus a business-driven activity that involves the management of processes in a business enterprise. Conversely, SOA is an IT-driven activity that facilitates the interoperability of services and business processes. Since Business process modeling provides a common and comprehensive understanding of a business process, they are essential in enhancing integration within an enterprise.

Web Services Business Models Supported in SOA Environment

Some of the most useful web service business models that are supported in a SOA environment include Data Flow Diagrams (DFD) and Role activity diagrams (RAD). Data Flow Diagrams highlight how data flows between different places within a system. They offer an illustration of the various processes, showing how processes link together between each other, the users and outside wide through data stores. Aguilar-Saven (2004) states that by using a DFD, a systems analyst has the ability to specify processes at the logical level, and therefore showing how information gets in and leaves a process and where data will be stored within a process. Therefore, such a web service model can be supported within SOA environment during application integration at the business logic level. Role activity diagrams (RAD) are also other alternative web services business models that are supported in SOA environment. They present a graphic view of a process based on individual roles and how these roles interact with each other. Since RADs can demonstrate how processes interact, they can be supported in SOA environment to illustrate how communication and applications will interact.

SOA And BPM Work Together

Though SOA and BPM are two independent and distinct measures, they can both be implemented within an enterprise to offer a platform that aligns and integrates business and IT around business processes. As highlighted by Aguilar-Saven (2004), both BPM techniques and SOA tools cannot in themselves offer 'the solution' to a business. However, they aid business analysts to effectively design and manage processes that drive communication and consensus in a firm. Since BPM is a strategy that manages the performance of a business through optimizing business processes, it can prove to be costly and complicated to manage various independent business processes that span in multiple applications. Conversely, though SOA supports the creation of reusable services to improve the integration of applications, such services cannot be self-optimize. However, adopting both SOA and BPM can be leveraged to offer a platform that aligns and integrates business, and IT functions around a business process. Therefore, both BPM and SOA can be adopted together within the FESL Company, where BPM will define business processes at a higher abstraction level, and SOA will provide services to support and interoperate these processes.

Modelling Tools to Transition from BPM To SOA

Some of the modeling tools that are required to transform from BPM to SOA include ARIS Platform and BIMM. The Business-IT-Mapping Model (BIMM) describes the way through which activities existing in the business process are transformed into activities of the system process. In addition, all system process activities can also be traced backward to the business process activities from where they originated from. More specifically, the main function of BIMM is to highlight existing relationships between business processes and system processes. Since it does not describe the control flow between activities of the system process model, only individual activities have to be imported into the system process modeling tool. Additionally, the ARIS Implementation Platform enables the design of SOA based on the business processes of an organization. Analysts can be able to use the business process designs to determine the necessary performance of individual services, and how these services have been orchestrated to make up a process. Therefore, in the case of FESL, ARIS and BIMM modeling tools would be essential to transition to SOA and control dependencies.

Task 3

IBM

IBM's SAP enterprise web service business strategy approach is to provide a centralized system for enhanced decision making, reducing costs and extending business efficiency. Their web services allow the integration of an organization's set of heterogeneous applications. Additionally, IBM supports interoperability of different applications through the company's existing systems with new SAP systems. In retrospect, this minimizes the dependencies between components of an application. In cases where one component fails, replacement and independent development can be easy to undertake without interfering with the overall performance of a system. Since SAP systems adopt open standards, companies can be able to reduce the chances of vendor lock-in, and also expanding the scope of tools that can be connected to the overall system (IBM, 2015). Business can, therefore, be able to maintain a comprehensive, accurate and securely integrated system that reduces costs associated with upgrading and maintenance of existing applications. In situations where businesses need to move large volumes of data between various systems, SAP systems extract data from the source and transform it into a common standard before being loaded to the destination. Therefore, businesses can gain a competitive advantage by avoiding errors, reducing the time and maximizing the quality of data.

Amazon Web Services

The business strategy behind the deployment of Amazon Web Services is minimizing costs and complexities through ensuring dynamic scalability of storage capacity. A common characteristic in most AWS deployment services is the ability to improve the flexibility of enterprise applications through automating these services. This strategy ensures that firms that want to start small and scale up with time can handle traffic and large data without incurring additional costs of purchasing and maintaining physical infrastructure (AWS, 2015). Since most companies experience high peaks of workload for short periods of time, Amazon's web services ensure that businesses can scale their computing capacities by increasing or decreasing workloads dynamically. Additionally, businesses are also able to monitor the rate of resource utilization to make faster and thus make informed decisions of when to scale up or down. Accordingly, the deployment of AWS is to enhance the scalability of storage and computing capabilities, where business creates more value from reduced costs and efficiency associated with the Amazon's web services. The company also offers a wide range of services for companies to select from, which allows companies to select packages that are consistent with their requirements.

The Microsoft Azure

The business strategy behind the deployment of Microsoft Azure is to enhance collaboration and thus improve productivity, sales, ensure quality customer service and foster innovation. In the "City of Hope" case study, the medical research team collaborates with other members using Microsoft's web services to carry out innovative research, aggregate data and offer education courses (Microsoft Azure, 2017). This speeds up their speed and efficiency in developing drugs, and henceforth able to overcome the challenge of analyzing robust and distributed data. Brainshark Company also utilizes web services to drive sales and ensure efficient customer care through training staff spread at different geographic locations. This has been effective for the company, where the CEO claims that no expenses have been spent on purchasing hardware. Another business strategy evidenced from the deployment of Microsoft Azure is based on improving the productivity of companies like Rolls-Royce and Real Madrid football club. By providing a common platform for staff, Rolls-Royce adopts analytical tools to filter big data and aggregate it to add its value. Real Madrid can create a platform to engage with fans and understand their needs. The aggregated data is used to increase revenue through implementing targeted marketing campaigns.

References

Aguilar-Saven, R. S. (2004). Business process modeling: Review and framework. International Journal of Production economics, 90(2), 129-149.

Ahalt, M. D. H. S. C., & Krishnamurthy, A. K. (2002). A Comparison of Java RMI, CORBA, and Web Services Technologies for Distributed SIP Applications. Industrial Informatics.

AWS. (2015). Overview of Deployment Options on AWS. Amazon Web Services, Inc. Retrieved from https://d0.awsstatic.com/whitepapers/overview-of-deployment-options-on-aws.pdf

Dumitrache, M., Dumitra, S., & Baciu, M. (2010). Web services integration with distributed applications. Journal of Applied Quantitative Methods, 5(2), 223-233.

He, W., & Da Xu, L. (2014). Integration of distributed enterprise applications: A survey. IEEE Transactions. Industrial Informatics, 10(1), 35-42.

Jestratjew, A., & Kwiecien, A. (2013). Performance of HTTP protocol in networked control systems. IEEE Transactions on Industrial Informatics, 9(1), 271-276.

IBM. (2015). IBM Software for SAP Solutions. International Business Machines Corporation. Retrieved from http://www.redbooks.ibm.com/abstracts/sg248230.html?Open

Khadka, R. (2013). Migrating a large scale legacy application to SOA: Challenges and lessons learned. Paper presented at the Proceedings - Working Conference on Reverse Engineering, WCRE, 425-432

Microsoft Azure (2017). Check out these innovative stories by world class companies. Microsoft Inc. Retrieved from https://azure.microsoft.com/en-us/case-studies/

Simanta, S., Morris, E., Balasubramaniam, S., Davenport, J., & Smith, D. (2009). Information assurance challenges and strategies for securing SOA environments and web services. Systems Conference, 2009 3rd Annual IEEE.

Versteeg, J. (2008). A method to leverage legacy Oracle Forms applications in an SOA. Utrecht, Netherlands: University of Twente.