Survey: Utilizing excess cloud resources to reduce the cost and time of distributed processing
Ali Jassim Sanad, Alauddin Yousif Al-Omary · 2019
Distributed computing requires to have large resources in order to perform. The distributed computing system performance is proportionally increasing along with the increasing of the available resources. A work obtained from a single computer in one hundred hours can be achieved within one hour by integrating a hundred computers together. This makes the distributed computing a preferred choice for applications that need to work on big data or intensive parallel processing. Maintaining a high number of resources is considered a costly issue, especially for the applications that operate in an infrequent period. On the other hand, the cloud service providers invested in building their data centers and preparing them to handle dynamically changing loads. By adding extra resources, the cloud providers successfully provide scalability features to their customers. Such excess resources can be leased with a relatively low cost compared with other services, but at the same time, it comes with some risks. Deploying the distributed computing on the cloud using the excess resources will help in solving many issues. From one point, it will help to minimize the total costs of provisioning which the distributed computing suffering from. Another point is that the cloud providers maintain large resources which can be used to make the process faster, while relaying on the distributed computing system to manage the risks associated with the excess resources allocation. In this paper, a survey of implementing distributed computing integrated with cloud is introduced. The paper shows some benefits and problems associated with such integration. The survey shows that utilizing the cloud excess recourses can reduce the total provisioning cost by 90% or more, but it has to be done using an enhanced version which is designed to tolerate the dynamic cloud changing.