Suhardi | Search arXiv

arXiv2021

X-Driven Methodologies for SOA System Development -- A Survey

This study aims to evaluate four service-oriented architecture (SOA) system software development methodologies: business-driven development, model-driven development, event-driven development, and domain-driven development. These methods, generically labelled as x-driven methodologies (XDMs), are commonly used in a general software development context, but software architects can also apply them in an SOA-based system. Each XDM typically focus on a specific aspect that drives its processes and steps. This aspect is indicated by its label. An evaluation method called qualitative screening mode is used in this study. XDMs are analysed based on their features to determine the suitability or support for service-oriented solutions. Criteria used to appraise each method are taken from SOA characteristics and SOA manifesto points. Of the four discussed XDMs, business-driven development is the best-suited approach to implement a service-oriented system shown by its conformity with the selected assessment criteria. Nevertheless, the other three XDMs have also their own strengths. Model-driven development is excellent for productivity, event-driven development is preferential for a quick response and asynchronous work, while domain-driven development is distinctive to describe problems precisely. The originality of this research is in the assessment general software development approaches of XDMs to be applied to SOA approach. The results can help developers in considering suitable methods to construct a prospective software system. Previous studies only investigate on methodologies designed intentionally for service-oriented systems.

cs.SE↗

arXiv2012

Performance Measurement of Cloud Computing Services

Cloud computing today has now been growing as new technologies and new business models. In distributed technology perspective, cloud computing most like client-server services like web-based or web-service but it used virtual resources to execute. Currently, cloud computing relies on the use of an elastic virtual machine and the use of network for data exchange. We conduct an experimental setup to measure the quality of service received by cloud computing customers. Experimental setup done by creating a HTTP service that runs in the cloud computing infrastructure. We interest to know about the impact of increasing the number of users on the average quality received by users. The qualities received by user measured within two parameters consist of average response times and the number of requests time out. Experimental results of this study show that increasing the number of users has increased the average response time. Similarly, the number of request time out increasing with increasing number of users. It means that the qualities of service received by user are decreasing also. We found that the impact of the number of users on the quality of service is no longer in linear trend. The results of this study can be used as a reference model for the network operator in performing services in which a certain number of users in order to obtain optimal quality services.

cs.NI↗