The History Of Distributed System

Published Date: 02 Nov 2017

Distributed System

[Student’s name]

[Course title]

[Supervisor’s name]

[Date]

Distributed System

Introduction

The paper aims to identify the approaches and methods that are used in the distributed system and examines the difference between the centralized systems against distributed system. The paper further attempts to describe the difficulties and technical issues associated with the distributed services. Lastly the paper will help to identify the significance of distributed system and helps in analyzing the benefits that are associated with the particular system.

Methods and Approaches in Distributed System

There are many approaches that are utilized in the forming of Distributed system in which some of the methods are described below:

Test Driven Development (TDD) is the software development process which helps in building up of the distributed system. It performs the testing of all the software and components associated with the distributed system. If any of the features are added in to the software, then the TDD test will be performed that will test the failure or the approval of those features (Vidyarthi, Sarker, Tripathi, & Yang, 2009). The software developers controls the entire process of testing in which the code is written by the user which will allow the test to pass or fail. The tests are the requirements for the approval of the feature because it is essential for the success of the entire system (Ghosh, 2010). The test contains many advantages for the system and the user. The main advantage is that the system will have higher chances of success and lower chances of failure. The new feature will then be approved and highly tested under the conditions required for the particular test. The behavior of the feature is then monitored by the software developers which they make are quite safe to use (Tanenbaum & Steen, 2010).

The TDD tests are more different in nature than the integrated test because test is conducted on the unit basis for the TDD. It is the kind of test in which the test is conducted on a single unit or a function rather than the whole software (Vidyarthi, Sarker, Tripathi, & Yang, 2009). The particular test can be run in isolated conditions from other components of the software or program. Many automates tests are run upon the system which makes the program even quicker which results in better performance and quality (Kamalapur & Deshpande, 2007).

Difference between Centralized System and Distributed System

The concept of a distributed system ensures that many independent computers are represented as one coherent system. The main examples of the distributed system include World Wide Web and internet. The central rack location is not identified in the distributed system. The concept is very much wise in the small projects but it has some weaknesses in the medium and large systems. Re-tuning and maintenance of the distributed system for larger projects is difficult (Veríssimo & Rodrigues, 2004). The term is more responsive if there is high difference in the nature of data gathering and the storage. A better price is offered by the distribution system that the mainframes offered by the centralized system. In the distributed system, if one of the component or machine crashes, then the whole system can still be running which is more efficient than the centralized system (Vidyarthi, Sarker, Tripathi, & Yang, 2009). In the centralized system, the system crashes if the server crashes. Therefore, the distributed system has a comparative advantage over the centralized system (Wu, 1999).

There is low cost and higher performance ratio for the distributed system, whereas, high cost and low performance ratio is associated with the centralized system. The computing power for a distributed system exceeds the computing power for mainframes offered by the centralized system. In the distributed system, the load is distributed to many computers due to which the speed of the system is enhanced, whereas, in the centralized system the load is gathered in one computer which increases its load and decreases the speed for the system (Vidyarthi, Sarker, Tripathi, & Yang, 2009). The computing power is also distributed in to small components in the distributed system which increases modular expandability. However, in the centralized system the computing power is not distributed in to a small component which decreases the modular expandability (Coulouris, 2005).

The distributed system allows the access of many users to remain beneficial and use the data base. The centralized system does not allow this access because the system does not have the potential. The resources are shared to other users in the distributed system which is impossible to achieve in other systems (Vidyarthi, Sarker, Tripathi, & Yang, 2009). The centralized system is efficient but not as efficient as the distributed system. Distributed system is more responsive and accurate according to the needs and desires of users, whereas, the centralized system is not much responsive. This remains very much practical for companies that are huge in size which requires many users to use the system in one time (Krakowiak & Shrivastava, 1991).

However, the centralized systems also have certain amount of advantages that cannot be overruled. In centralized system, the listing of the data is controlled by one user so the system is free from any inconsistencies and it has all the control of the data. A standard format shall be run which is influenced by the centralized system (Vidyarthi, Sarker, Tripathi, & Yang, 2009). The companies that have smaller size are mostly using the centralized system because they do not have much reach in the outside world. However, the major companies like IBM and General Motors which are large size uses the distributed system in order to stay flexible and maintain competitiveness in the market. Therefore, the centralized system is less efficient than the distributed system (Rai & Agrawal, 1990).

Difficulties and Technical Issues

There are many difficulties and technical issues associated with the distributed system. The main issue arises when the components are untrusted and unreliable. The design issues can also arise in the distributed system. Amongst these design issues, the setup of the model and identifying the components can become an issue (Krakowiak & Shrivastava, 1991). The interaction amongst components and its communication can also make faults. The wide system security is required for the system. This is the challenge which is mostly faced by the distributed system (Vidyarthi, Sarker, Tripathi, & Yang, 2009). There can be many challenges in the distributed systems which are discussed below:

Heterogeneity

The diversification of the distributed system includes the software and the hardware of the particular system. The heterogeneity in the scenario is described as the differentiable components of the system (Vidyarthi, Sarker, Tripathi, & Yang, 2009). Heterogeneity also includes hardware devices, operating systems, local networks and the language of the program. There might occur problems in this hardware and software’s which can cause technical issues (Crichlow, 2000).

Transparency

The complexity in the distributed system can also cause technical issues because it becomes difficult for the software developers to understand the particular setup of the software. The subtractions layer for the distributed system can be effective in the system which is absent which can also cause likely problems (Vidyarthi, Sarker, Tripathi, & Yang, 2009). Transparency is an effective component in the distributed system because a user that is using the internet does not know that when they search anything on the internet, their material comes in compressor before the search engine delivers them the result. There are many terminals and points in the transparency process which can likely to cause technical issues (Misra, 2010).

Openness

If the system does not have the well-defined interfaces then it would become difficult for the developers to add new features in to the system which will create problems. The social networking websites such as facebook and twitter provides permission to the software developers to interactively develop their software. Therefore, it is important that these specifications are met in the distributed system (Kamalapur & Deshpande, 2007).

Concurrency

The concurrency will not be maximized if there is no anticipation of the resource handling components. These are required for the assessment of the users. The task is considered to be very tricky because sometimes the system becomes unstable when the user tries to update or look in to the data. The distribution system is controlled by a multi user (Kamalapur & Deshpande, 2007).

Security

The security measurement of the distributed system should be strong to avoid any kind of outside invasions and to protect the system from intruders. Therefore, a strong security measurement system is required which can create difficulties for the user. Secure mechanism must be in order because sensitive information is available on the internet (Krakowiak & Shrivastava, 1991).

Scalability

The distributed system shall be scalable because the system can get hanged if multiple users are using the system. If the additions of users are handled by the system then it is said to be scalable. There are 3 dimensions of the distributed system namely the size, the geography, and the administration. Size refers to the number of users associated with the system, and the technical issue that can arise in the system related to the specific dimension is overloading of users in the system (Kamalapur & Deshpande, 2007). The second dimension is the geography in which the distance between the user and the resource is measured. The problem that arises in the particular dimension is reliability of its communication. The third dimension is the administration in which the controllability of the size of the distributed system is monitored. Therefore, problems can occur if there is controllability of the size or the users (Kamalapur & Deshpande, 2007).

Resilience to failure

Many components are associated with the distributed system due to which the chance of failure increases with the addition of features and components. If any or total components fail to work then the whole system will become a failure (Vidyarthi, Sarker, Tripathi, & Yang, 2009).

Conclusion

There are many concepts and methods which help in forming the distributed system. Amongst these many approaches, Test Driven Development and Unit test is ranked at top. Test Driven Development (TDD) is the software development process which helps in building up of the distributed system. It performs the testing of all the software and components associated with the distributed system. If any of the features are added in to the software, then the TDD test will be performed that will test the failure or the approval of those features. The concept of a distributed system ensures that many independent computers are represented as one coherent system. The main examples of the distributed system include World Wide Web and internet. The central rack location is not identified in the distributed system. The concept is very much wise in the small projects but it has some weaknesses in the medium and large systems. Re-tuning and maintenance of the distributed system for larger projects is difficult. The distributed system allows the access of many users to remain beneficial and use the data base. The centralized system does not allow this access because the system does not have the potential. The resources are shared to other users in the distributed system which is impossible to achieve in other systems. Amongst many technical issues in the distributed system, the setup of the model and identifying the components can become an issue. The interaction amongst components and its communication can also make faults. The diversification of the distributed system includes the software and the hardware of the particular system. The heterogeneity in the scenario is described as the differentiable components of the system.

The complexity in the distributed system can also cause technical issues because it becomes difficult for the software developers to understand the particular setup of the software. The subtractions layer for the distributed system can be effective in the system which is absent which can also cause likely problems. If the system does not have the well-defined interfaces then it would become difficult for the developers to add new features in to the system which will create problems. The concurrency will not be maximized if there is no anticipation of the resource handling components. These are required for the assessment of the users. The security measurement of the distributed system should be strong to avoid any kind of outside invasions and to protect the system from intruders. The distributed system shall be scalable because the system can get hanged if multiple users are using the system. If the additions of users are handled by the system then it is said to be scalable. There are 3 dimensions of the distributed system namely the size, the geography, and the administration. Many components are associated with the distributed system due to which the chance of failure increases with the addition of features and components. These are the difficulties which are mostly faced in distributed systems.