Fairness-Aware Task Allocation for Heterogeneous Multi-Cloud Systems

Abstract

Cloud computing is rapidly growing for its on-demand services over the Internet. The customers can use these services by placing the requirements in the form of leases. In IaaS cloud, the customer submits the leases in one of the form, namely advance reservation (AR) and best effort (BE). The AR lease has higher priority over the BE lease. Hence, it can preempt the BE lease. It results in starvation among the BE leases and is unfair to the BE leases. In this chapter, the authors present fairness-aware task allocation (FATA) algorithm for heterogeneous multi-cloud systems, which aims to provide fairness among AR and BE leases. We have performed rigorous experiments on some benchmark and synthetic datasets. The performance is measured in terms of two metrics, namely makespan and average cloud utilization. The experimental result shows the superiority of the proposed algorithm over the existing algorithm.

Time-Aware Task Allocation for Cloud Computing Environment

“…The objective of this problem is to achieve fairness among the leases by assigning the leases in the precise order. To the best of our knowledge, this problem is not well-covered in the recent literatures (Li et al, 2012;Panda, Pradhan, Neha, & Sathua, 2015).…”

“…Note that the fairness time is the product of a threshold value (i.e., (0 ~ 1]) and the execution time of the executed AR lease. The algorithm focuses on the time duration of the leases rather than the number of leases to achieve fairness as used in (Panda, Pradhan, Neha, & Sathua, 2015). We conduct simulations on the proposed algorithm using synthetic datasets.…”

“…However, frequent postpone of BE lease leads to starvation and dissatisfaction for the BE customer. Therefore, lease allocation algorithms (Li et al, 2012;Panda, Pradhan, Neha, & Sathua, 2015) must incorporate these issues for the efficient assignment of leases. Many practitioners have proposed fairness algorithms to provide better services for both AR and BE leases.…”

“…Recently Panda, Pradhan, Neha and Sathua (2015) have proposed fairness-aware task allocation algorithm for heterogeneous multi-cloud systems. The algorithm permits few BE leases to execute before AR leases in a regular time interval.…”

An Efficient Intra-Server and Inter-Server Load Balancing Algorithm for Internet Distributed Systems

“…These algorithms calculate the minimum completion time, which is the sum of execution time and ready time to balance the workloads. Panda, Neha and Sathua (2015) have considered the uncertainty nature of resources in cloud task scheduling and proposed an uncertainty-based task scheduling algorithm. They have shown the performance of the algorithm in terms of makespan, uncertainty, save makespan, delay makespan and average cloud utilization.…”