Fault tolerant workflow scheduling based on replication and resubmission of tasks in Cloud Computing

Journal Title: International Journal on Computer Science and Engineering - Year 2012, Vol 4, Issue 6

Abstract

The aim of workflow scheduling system is to schedule the workflows within the user given deadline to achieve a good success rate. Workflow is a set of tasks processed in a predefined order based on its data and control dependency. Scheduling these workflows in a computing environment, like cloud environment, is an NP-Complete problem and it becomes more challenging when failures of tasks are considered. To overcome these failures, the workflow scheduling system should be fault tolerant. In this paper, the proposed Fault Tolerant Workflow Scheduling algorithm (FTWS) provides fault tolerance by using replication and resubmission of tasks based on priority of the tasks. The replication of tasks depends on a heuristic metric which is calculated by finding the tradeoff between the replication factor and resubmission factor. The heuristic metric is considered because replication alone may lead to resource wastage and resubmission alone may increase makespan. Tasks are prioritized based on the criticality of the task which is calculated by using parameters like out degree, earliest deadline and high resubmission impact. Priority helps in meeting the deadline of a task and thereby reducing wastage of resources. FTWS schedules workflows within a deadline even in the presence of failures without using any history of information. The experiments were conducted in a simulated cloud environment by scheduling workflows in the presence of failures which are generated randomly. The experimental results of the proposed work demonstrate the effective success rate in-spite of various failures.

Authors and Affiliations

Jayadivya S K , Jaya Nirmala S , Mary Saira Bhanu S

Keywords

Related Articles

Solving Sparse Rating Problem Using Fine Grained Approach

Recommender System is a system that automatically recommends all similar kind of items that are of user interest. In design of the recommender systems rating is the crucial issue. Till today many algorithms have been pro...

ARPE : An Attack-Resilient and Power Efficient Multihop WiMAX Network

IEEE 802.16 standard known as WiMAX (Worldwide Interoperability for Microwave Access), is one of the most promising wireless access technology for next generation all-IP networks. The fundamental requirements for WiMAX...

SECURE ROUTING IN WIRELESS SENSOR NETWORKS: ROUTING PROTOCOLS

A Secure Routing capabilities on the order of Node to Base Station will be capable of both wired connectivity to the internet as well as wireless connectivity to the sensor network. One aspects of sensor networks organiz...

STUDY OF THE BEHAVIOR MODELS BASED ON PROBABILITY AND TIME BY USING MARKOV PROCESS AND TRANSITION MATRIX.

This paper suggests an approach to software system architecture specification based on behavior models. The behavior of the system is defined as an event occurring probability and time. The behavior of the system over a...

Spam Classification using new kernel function in Support Vector Machine

Due to the increase in internet users, there is a rapid growth in spam e-mails. In recent years, kernel function have received major attention, particularly due to the increased popularity of Support Vector Machine. It i...

Download PDF file
  • EP ID EP156014
  • DOI -
  • Views 93
  • Downloads 0

How To Cite

Jayadivya S K, Jaya Nirmala S, Mary Saira Bhanu S (2012). Fault tolerant workflow scheduling based on replication and resubmission of tasks in Cloud Computing. International Journal on Computer Science and Engineering, 4(6), 996-1006. https://europub.co.uk/articles/-A-156014