Title: Dissection of the experimental outcome of split-protocol

Authors: Bharat S. Rawal; Qiang Duan; Pandi Vijayakumar

Addresses: Department Information Sciences and Technology, Penn State Abington, Abington, PA 19001, USA ' Department Information Sciences and Technology, Penn State Abington, Abington, PA 19001, USA ' University College of Engineering Tindivanam, Melpakkam, Tindivanam 604 001, India

Abstract: The split-protocol theory was developed for load balancing and quicker data communication. Split-protocol computing paradigm uses web services on geographically distributed web servers on the cloud. A system of large split-servers that form the cloud to handle computing and storage task that would otherwise create a massive CPU utilisation if we work with the traditional individual server. In an earlier paper, we established that the application split-protocol produces higher performance compared to traditional clusters. Based on the necessity, diverse types of split configurations are applied for higher throughput, better response and connection time. The experiential throughput enhancement was within the range 6.5%-25% over non-split systems. This paper examines empirical results split systems to understanding its behaviour compared with non-split systems. Split-protocol was implemented on private cloud for internal data servers of the organisation, not made available to the general public. The split concept emerged from the HTTP/TCP/IP network protocol implementation. The split-system model with given sets of constraints can produce better throughput than conventional equivalent server systems. In this paper, we have presented the analytical model to support a high performance of split-protocol implementation. We have also mathematically evaluated the inherent reliability characteristics of the split-system.

Keywords: parallel computing; performance; pipeline; splitting http requests; cloud computing.

DOI: 10.1504/IJAIP.2018.089487

International Journal of Advanced Intelligence Paradigms, 2018 Vol.10 No.1/2, pp.23 - 44

Received: 04 Feb 2016
Accepted: 06 Apr 2016

Published online: 29 Jan 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article