Parallel queue scheduling in Dynamic Cloud environment using Backfilling algorithm

Title

Parallel queue scheduling in Dynamic Cloud environment using Backfilling algorithm

Creator

Jayapandian N.

Description

Cloud Computing reshapes the entire computing paradigm. In general, cloud computing means outsourcing available services and data storage in centralized scenario. In cloud computing task allocation is a major problem because multiple numbers of tasks are allocated to multiple numbers of processors for simultaneous processing. From the given list, tasks are queued according to the ascending order based on their duration. This paper is designed to solve the Task Scheduling problem, by using our proposed effective new approach of Backfilling algorithm. Depending upon the task duration, tasks are split into multiple threads for processing. Multiple thread tasks are processed in the basic concept of "gang scheduling" technique. Here we implement new backfilling algorithm concept to minimize the idle processing time of the processors. The existing Simple Backfilling Algorithm (SBA) is used to minimize the ideal time processing. Whereas comparatively Dynamic Cloud Scheduling using Backfilling Algorithm (DCBA) is designed to reduce the ideal time processing than SBA to carry out the process of both LQueue and SQueue simultaneously. At the outset, DCBA reduces the average waiting time. As mentioned the algorithm which is specified in the previous line that contains three level which represent the working speed of the algorithm. The first and second level of DCBA algorithm is comparatively similar to the performance of SBA algorithm. The maximum better performance was given in a queue size (q=1.5) by DCBA algorithm as compare to SBA algorithm. The existing type (Gang Scheduling) consist of two approaches namely Adaptive First Come First Serve (AFCFS) and Largest Job First Served (LJFS) that focus on non-parallel jobs with deadline. When compare to existing gang scheduling algorithm and SBA algorithm the average waiting time of DCBA has slight improvement in the loader level of the key. As the separation of the queue like LQ and SQ the waiting time and average waiting time is reduced comparatively. 2018, Intelligent Network and Systems Society.

Source

International Journal of Intelligent Engineering and Systems, Vol-11, No. 2, pp. 39-48.

Date

2018-01-01

Publisher

Intelligent Network and Systems Society

Subject

Backfilling; Cloud computing; Dynamic cloud; Gang scheduling; Grid computing; Largest queue; Load balancing; Shortest queue; Task scheduling; Thread

Coverage

Jayapandian N., Christ University Faculty of Engineering, Bangalore, India

Rights

All Open Access; Bronze Open Access

Relation

ISSN: 2185310X

Format

Online

Language

English

Type

Article

Identifier

https://doi.org/10.22266/ijies2018.0430.05

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85050876104&doi=10.22266%2fijies2018.0430.05&partnerID=40&md5=fdfd77618b9419adbfdc2e1a77530924