energy-aware task scheduling using aco in cloud

Post on 27-Dec-2015

88 Views

Category:

Documents

1 Downloads

Preview:

Click to see full reader

DESCRIPTION

Energy-aware Task Scheduling using Ant Colony Optimization in Cloud

TRANSCRIPT

Linda . J, Ananthanarayana V.S.NITK Surathkal

Energy Aware Task Scheduling using Ant Colony Optimization in Cloud

•Introduction •Need for Task Scheduling & Energy Awareness•Problem Statement & Objective•Proposed Solution & Methodology•Results•Conclusion

Agenda

•“BURST” (Sudden increase or decrease) natured Web application demands affected the business.•Cloud Computing provided the solution via different service models like,•IaaS, PaaS, SaaS•Deployment Models Public, Private & Hybrid Cloud•Cloud computing is a model for enabling ubiquitous, convenient, on-demand access to a shared pool of resources.

Cloud Computing

•Many users access Virtual Machines everyday. •Efficient Task Scheduling Algorithms are required to increase profit for the cloud providers.•Consequently, Servers are always ON thereby increasing the Total Energy Consumption.•There arises a need to reduce the energy consumption in datacenters.

Need for Energy Awareness in Task Scheduling

•Scheduling [3] the n tasks (T1, T2,…,Tn) to m Virtual Machines (VM1, VM2,…,VMm) running on p Physical hosts (P1, P2,…,Pp) in such a way that maximum completion time or makespan of these n tasks will be minimized.•n>m>p

Task Scheduling

•To go through all possible (Task,VM) pairs so as to reduce the makespan.•To go through the Total Energy Consumption in all the hosts.

Requirements of Energy-aware Scheduler

•The Problem is to design a Energy-aware Task Scheduling Algorithm.

•Objective:–To design a Task scheduling Algorithm.–To add the Energy Awareness factor to the technique.–To compare the Energy-aware Algorithm with existing algorithms.

Problem Statement

•Ant Colony Optimization for Task Scheduling

Proposed Solution

System Model

•Cloudsim [16]•CIS registry hold information about the resources•Scheduler (or Broker) is enhanced to be an energy-aware scheduler

System Model

•LP Model (Based on [13])

•ϕ = CPU Utilization•Pidle = Power when CPU is idle•Pmax = Power when CPU is fully utilized

•RT Model (Based on [12] )

•The Expected Time to Compute is given by

•Wi = Workload •CCj = Computing Capacity

Initial Pheromone

Methodology

Task Rule•An ant randomly samples a task node from the list of task nodes yet to visit J’k

VM Rule • An ant k positioned at task node r, selects a VM node s, by

η(r,s) = inverse of makespan till s, 1/ CTmax

Vk = list of visited VM nodes

Θ(v) = Completion time of last Job in v

β is a parameter which determines the importance of pheromone.

Methodology contd.

Global Updation Rule•Once all ants have built their tours, pheromone is updated on all edges according to,

α is a pheromone decay parameter. Δτ(r,s)= 1/Lk , Lk is the length of the tour performed by ant k and m is the no. of ants.

Local Updation Rule•While building a solution (i.e., a tour), ants visit edges and change their pheromone level by applying the local updating rule shown below.

Methodology contd.

VM Rule

Where ω(r,s)= inverse of total power consumed in hostsμ(h) = power consumed in host h

γ is a parameter which determines the importance of power consumption

Energy-awareness

Algorithm

Results

•22% improvement over FCFS

Energy in

kWh

Job Mix

•In this project, a new task scheduling algorithm

using Ant-colony optimization that reduces the

power consumption for cloud is proposed.

•The proposed method outperforms the existing

method by 22% under the experimental conditions.

Conclusion

1. Kun Li, Gaochao Xu, Guangyu Zhao, Yushuang Dong, Dan Wang, Cloud Task schedul- ing based on Load Balancing Ant

Colony Optimization, IEEE, 2011.

2. Marco Dorigo, Luca Maria Gambardella, Ant Colony System: A Cooperative Learning Approach to the Travelling Salesman

Problem, IEEE Transactions, April 1997.

3. GU Srikanth, VU Maheswari, AP Shanthi, A Siromoney, Tasks Scheduling Using Ant Colony Optimization, Journal of

Computer Science, 2012

4. Alberto Colorni, Marco Dorigo, Vittorio Maniezzo, Marco Trubian, Ant System for Job-shop Scheduling, Belgian Journal of

Operations Research, 1994.

5. Mohsen Amini Salehi, P. Radha Krishna, Krishnamurty Sai Deepak and Rajkumar Buyya, Preemption-aware Energy

Management in Virtualized DataCenters, IEEE, 2012.

6. Ying Chang-tian, Yi Juong, Energy Aware Task Scheduling using Genetic Algorithms, IEEE, 2012.

7. Eugen Feller, Louis Rilling, Christine Morin, Energy-Aware Ant Colony Based Work- load Placement in Clouds, INRIA,

IEEE/ACM Conference on Grid Computing, May 2011.

8. ODC Alliance Carbon Footprints Values

9. Rodrigo N. Calheiros, Rajiv Ranjan, Anton Beloglazov, C ́esar A. F. De Rose and Rajkumar Buyya, CloudSim: a toolkit for

modeling and simulation of cloud computing environments and evaluation of resource provisioning algorithms, IEEE, 2010.

10. JimBlythe, SonalJain, EwaDeelman, Yolanda Gil, KaranVahi, TaskSchedulingStrategies for Workflow-based Applications in

Grids, IEEE, 2005.

11. A. Belaglazov and R. Buyya, “Optimal Online deterministic algorithms and adapative heuristics for energy and performance

efficient dynamic consolidation of Virtual Machines in Cloud Datacenters”, Concurrency and Computation: Practice and

Experience, 2011.

12. Ali, S., Siegel, H.J., Maheswaran, M., and Hensgen, D.: “Task execution time modeling for heterogeneous computing

systems”, Proceedings of Heterogeneous Computing Workshop, pp. 185–199, 2000.

13. T. Guerot, Thiery Monteil, Georges Da Costa, Rodrigo Neves Calheiros, Rajkumar Buyya, Mihai Alexandro, Energy-Aware

Simulation Using Dvfs,Simulation Modelling Practice And Theory, Elsevier 2013.

14. Josep Ll. Berral, à ñIgo Goiri, Ramã³N Nou, Ferran Juliã , Jordi Guitart, Ricard Gavaldã , Jordi Torres, Towards Energy-

Aware Scheduling In Data Centers Using Machine Learning, In Proceedings Of The First International Conference Oon

Energy-Efficient Computing And Networking, Acm 2010.

15. Armel Esnault, Eugen Feller, Christine Morin, Energy-Aware Distributed Ant Colony Based Vm Consolidation In Iaas

Cloud,Simulation Modelling Practice And Theory, Elsevier 2013.

16. Rodrigo N. Calheiros, Rajiv Ranjan, Anton Beloglazov, Cesar A. F. De Rose, and Rajkumar Buyya, “CloudSim: A Toolkit for

Modeling and Simulation of Cloud Computing Environments and Evaluation of Resource Provisioning Algorithms”, Software:

Practice and Experience, Volume 41, Number 1, Pages: 23-50, ISSN: 0038-0644, Wiley Press, New York, USA, January

2011.

17. http://www.energy.wsu.edu/Documents/Data%20Center%20Energy%20Savings_Feb2013.pdf at 2.19 pm IST, April 30, 2014

References

Thank You

top related