Abstract¬¬¬ Cloud computing is the hottest topic in today’s world and is being used by more IT companies due to its benefits in cost saving and its ease of use .But when data is shifted from one data centre to other data centre, huge amount of CO2 gases emits and its consumed more power. Since energy is an important issue that is why green cloud computing is generated. Green cloud computing provides techniques and algorithm by the use of that huge amount of CO2 gases can be reduced and also the power consumption that are damaging the environment. Cloud provides many facilities to its user such that sharing of resources for different purposes .As cloud is developing its facing challenges and one of them is scheduling .Job scheduling is one of the key issue in green cloud computing because it’s one of the main job to be performed in order to get highest profit. Scheduling can decrease the power consumption so many algorithms has been proposed for this purpose and a lot more had to be done. In these paper we surveyed about various energy efficient job scheduling algorithms in cloud computing that are used for scheduling the jobs in order to provide energy effciency .
Keywords Cloud computing, Green cloud computing, energy efficiency, job scheduling, scheduling algorithms.
I. INTRODUCTION
1.1 Cloud Computing
Cloud computing is an internet based computing the data center are practiced by using a network of remote servers that are hosted on the internet where storing, managing and processing of data is done rather than on any personal computer or any local server. Its provides users to share data and resources to any computers or devices on demand.
Cloud computing provides three different approaches to cloud services layer that are –
• Infrastructure as a services referred as IaaS
• Platforms as a services referred as PaaS.
• Software as a services referred as SaaS.
These three services layers helps in business organizations and government organisations which cut down the operational expenses. IaaS provides services where user has control over operating systems IaaS provides services like Amazon Web services, GoGrid servepath. PaaS provides platform for developing applications by the help of cloud and there is no need to install any platform if the user is using it on its own machine. PaaS service like Windows azure, Google App Engine etc, SaaS used for running the existing applications like Instagram, Facebook, Google Apps here user does not deal with any installation of any software on their personal computer.
1.2 Green cloud computing
Cloud contain thousands of data centres to fulfil the demands of customers online and on time and because of that data centres is span in an area of hundred to thousand feet. Very huge amount of power is required to run these server.
So, Green cloud computing[] is not only picturised to achieve the efficient processing and utilised the computing infrastructure but its also for minimizing the energy consumption that are effecting the environment in a larger amount. Studies shows that average utilization of data centres can be nearly 20% and energy consumed by the idle resources is can be as much as 60%[14]. Green cloud computing is extremely important for ensuring that the coming year of cloud computing must be sustainable and eco-friendly. Green cloud works on large scale companies like Google ,Amazon IBM cloud because its enables user to get services anywhere. Green cloud computing is the on demand services because its shares big resource pool that user can buy on request or requirement. In the paper[11] proposed by Bharti wadhwa and Amandeep verma they have reviewed about the various researchers and their strategies to make cloud computing more energy efficient, to reduce the carbon emission from the environment and found virtualization can help in utilization of resources in clouds.
1.3 Scheduling
Scheduling is one of the crucial activities that executes in the cloud computing environment. Scheduling increases the efficiency of the work load of cloud computing. It is the job to get maximum profit , throughput and make-span. Scheduling algorithms in cloud computing is to utilize the resources meanwhile balancing the load between the resources so that the resources execute in a minimal execution time [14]. There are many scheduling algorithms in cloud computing. Scheduling algorithms is done to obtain high performance, utilize the resources , managing the load between the resources. Examples of Scheduling Algorithms are (FCFS)First Come First Serve, Round Robin Algorithm, Shortest Job First(SJF),Min-Min algorithm, Max-Max Algorithm and so on.Green scheduling algorithms [10] that based on neural predictors based on neural predictors can saved up to 70% of power savings.These algorithms enable us to cut down the data centre energy costs and lead us to strong competiative cloud computing environment.
1.4 Job scheduling
In a single data centre thousands of servers runs at any instance of time and at the same time the cloud system continuously receiving the batches of task requests that to be executed. At these time where server is busy than one has to notice some target servers out of those servers which are powered on ,which can fulfil the batch of incoming jobs. That’s why job/task scheduling is an important issue which greatly effect the performance of cloud services. Job scheduling is categorised on the basis of load balancing, temperature based and energy efficiency.
1.5 Energy efficiency
The goal of energy efficiency is to reduce the amount of energy consumed by any organisation in order to achieve or provide user good products and services. Reducing energy reduces the cost and it can be seen as to reduce the greenhouse gases. As today’s world users are more on cloud computing where number of tasks are executed at a time, power of consumption is reduced and CO2 emission is increasing in alarming rate so energy efficient scheduling algorithm is required in cloud computing.
The remainder of the paper contains the following – section II gives the motivation of the research , section III contains the objective behind the survey on that topic, section IV contains the literature survey and section V presents the conclusion.
II. MOTIVATION
Cloud computing is an important model in Information Technology(IT) field where huge number of task are done and requirement of resources are more because of that servers becomes busier. In data centre, where all servers physical resources are available, machines are consuming more power and emits heat which is effecting the environmental conditions. This is due to companies give more focus to the higher availability than on energy efficiency. The main aim of Green cloud computing is to reduce the energy that are consumed by the physical resources in data centre and save energy and also increases the performance of the system. And the Technique that is used for reducing the power consumption and make algorithm energy efficient is job scheduling. Many work has been done in the field of job scheduling for increasing the throughput, performance but a lot more have to be done on energy efficiency.
III. OBJECTIVE
In today’s world user are more interested in cloud computing because its saving their cost and time but due to its increasing usage its effecting the environment. The main objective of the paper is to surveyed through all the scheduling algorithms that are made for energy efficiency and that reduces the CO2 emissions from the environment.
IV. LITERATURE SURVEY
The following job scheduling algorithms are currently present in clouds and these algorithms are summarized.
1) A Pre-emptive priority based job scheduling algorithm
In these paper [1] author has proposed a algorithm of energy efficient scheduling that focuses on the pre-emptive part as well as it calculates the energy consumption for scheduling the jobs on the cloud computing servers. The main aim of the author behind these paper was to minimize the CO2 gases and maximise the resources but according to the suitability of servers. In these paper the computing server is selected on the basis of that satisfies the minimum resource requirement on a job. Resources are allocated by the method of best allocation scheme that saves a energy consumption by creating a balance between the power consumption and work load on the servers.
2) Energy Efficient scheduling with traffic load balancing. In these paper the author [2] has proposed scheduling algorithm that optimizes the energy consumption of data centre equipment while provides load balancing of traffic that is flowing within the data centre. And effective distribution of network traffic improves QoS of running cloud application by delaying the congestion related packet losses. The author has used Green Cloud simulator for the experimental results on algorithms. The proposed algorithm shows result that there is no increase in the energy consumption for commonly paper management in data centres.
3) Priority Based job scheduling algorithm
In these paper [4] the author has proposed an algorithm that is two fold mechanism. The first phase calculates the priority of the task that is given according to the specialised attributes in the tasks and then it sorted by the calculated priority. Second phase computes the time that each task will require to execute. The paper presents an algorithm that schedule the task in an order that apart from calculating the priority of tasks it will calculate the expected executable time of different task on different servers. The author uses CloudSim simulator for the experimental results. And the priority is calculated under different attributes such as user level, expected priority, waiting time and based on the formula each attributes of priority can be achieved.
4) Shortest job first scheduling
In these paper [5] the author has proposed a approach to minimize the energy in a cloud computing which is totally based on enhancing the green scheduler which performs the work load consolidation on a minimum servers in green cloud computing by the help of executing task which have minimum arrival time. This paper presents a data centre scheduling approach that helps in reducing the power consuming and achieves balance between the two factors energy efficient and performance. The author has used green cloud simulator for the experimental results.
5) Greedy scheduling of tasks with time constraints
In these paper [6] the author aim was to find an optimum task scheduling scheme to minimize the task response time and energy consumed by the data centre servers. The authors uses the most efficient server first scheme where the server with the highest computing capacity will provide a lower energy expenditure per processed job then optimization problem can be interpreted as a greedy assignment scheme. It is considered that the central scheduler sorts the servers on the basis of their energy efficiency and its assigns the tasks first to the most energy efficient servers. The author uses MATLAB simulation.
6) Green scheduling algorithm for energy savings in cloud computing
In the paper [7] the author has proposed a green scheduling algorithm that concentrates workload on the servers and then turn it off. They uses the neural network based predictor for savings energy in cloud computing. The predictor can predict the future load demand based on historical demand. The green scheduling algorithm determines which server should be turn off/on. It will turn on when the work load is more and turn off when the work load is lesser. In these paper author uses CloudSim and GridSim toolkit .
7) Workflow job scheduling in Green Cloud computing
In these paper [8] the author has proposed an energy efficient scientific workflow scheduling algorithm that is multi-step heuristics workflow scheduling algorithm named EARES-D Energy Aware Resource Efficient Workflow Scheduling under deadline constraint .It addresses the various objectives that include guaranteed QoS, reduce energy and CO2 emissions for energy efficient and eco-friendly data centres. Their proposed algorithm aims to meet the response time requirement and minimize the VM virtual machine for reducing the energy consumption. The author has used Java based CloudSim toolkit for green cloud computing infrastructure and evaluate the scheduling algorithm. The simulation results of their proposed algorithm shows that energy consumption decreases up to 11% from e-ECT to DVFS, up to 12% from FWS to BWS and they able to achieve average 30% of energy savings and resource utilization rate increases from 20% to 25%.
8) Energy Aware scheduling of real time and non real
time tasks
In these paper the author [12] has proposed a
algorithm That handles real time and non real time
tasks. They Used three processor the first two
processor handle Algorithm that handles real time
and non real time tasks. They used three processor
the first two processor handle real time tasks using
earliest Deadline First (EDF) and Earliest
Deadline Late (EDL) scheduling algorithms and
also real-time tasks its uses a StandBy Sparing
Technique and for Non Real time tasks are
scheduled using FCFS First Come First Serve
scheduling algorithm. For simulation the authors
used MATLAB. The algorithm that they have
proposed for both tasks had conserved energy up to
58% when compared to any power management and
up to 4% energy when compared to existing system.
9) Efficient energy scheduling for workflow tasks of Hybrids and DVFS enabled cloud computing
In these paper [13] the author has proposed an DVFS enables energy efficient workflow task scheduling in which the energy aware method distributes the parallel applications to the optimal processors and deals with them to reduce the energy consumption on a appropriate time slots and meet the required performance to execute the workflow within a deadline. The evaluation is done between the DEWTS with two heuristics algorithms ,HEFT and EES..They merge processors that much to find out the processor which are relatively inefficient and reassign the tasks to the appropriate time slots and scale their execution time. The experiment was carried out by the use of CloudSim simulator. The experimental results shows that the when CCR is 0.5 the EES saves upto 44% of energy and when CCR is 2 it reduces 37% enrgy dissipation.And when CCR is 2 DEWTS saved energy upto 46%.
V. CONCLUSION
In these paper we surveyed the various existing energy efficient job scheduling algorithms in green cloud computing. There are many parameters that can be mentioned as a factor of scheduling problem that can be considered — such as load balancing, throughput, service cost and so forth. Some of the results that shows that these algorithm have proved to be energy efficient like in the workflow job scheduling algorithm have saved up to 30% of energy. In Energy Aware scheduling algorithms for real and non real time tasks had saved up to 58% of energy. In most of the algorithm authors have used toolskit for finding the experimental results they were CloudSim and MATLAB toolkits. In Shortest Job First Scheduling and in e-STAB the experimental was carried out by the use of Green cloud simulator. There are certain aspects that should be considered as topics of research to produce more accurate and improved algorithms.
References
[1] Gaganjot Kaur , sugandhi Midha , A preemptive Priority based job scheduling algorithm in green cloud computing ,2016 6th International Conference-clous system and Big Data engineering(confluence)
[2] Dzmitry Kliazovich, Sisay T. Arzo, Fabrizio Granelli, Pascal Bouvry and Samee Ullah Khan , e-STAB: Energy-Efficient Scheduling for Cloud Computing Applications with Traffic Load Balancing,
2013 IEEE International Conference on Green Computing and Communications and IEEE Internet of Things and IEEE Cyber,
Physical and Social Computing
[3] Kapil Kumar, Abhinav Hans, Ashish Sharma, Navdeep Singh,
Towards The Various Cloud Computing Scheduling Concerns: A Review,
[4] Pankajdeep Kaur1 and Parampreet Singh ,
Priority based Scheduling Algorithm with Fast Task Completion Rate in Cloud
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 2, ISSUE 11, NOVEMBER 2013
[5] Abeer H. El Bakely, Hesham A.Hefny , Using Shortest Job First Scheduling in Green Cloud Computing,
International Journal of Advanced Research 9, September 2015
[6] Ziqian Dong, Ning Liu and Roberto Rojas-Cessa Dong et al , Greedy scheduling of tasks with time constraints for energy-efficient cloud computing data centers,
Journal of Cloud Computing: Advances, Systems
and Applications (2015)
[7] Truong Vinh Truong Duy Yukinori Sato and Yasushi Inoguchi , Performance Evaluation of a Green Scheduling Algorithm for Energy Savings in Cloud Computing,
[8] Fei Cao, Michelle M. Zhu , Energy Efficient Workflow Job Scheduling for Green Cloud, 2013 IEEE
[9] Er. Shimpy, Mr. Jagandeep sidhu Different scheduling algorithms in different Cloud environment
International Journal of Advanced Research in Computer and Communication Engineering 9, September 2014
[10] B. Gayathri , Green Cloud Computing
Third International Conference on Sustainable Energy and Intelligent System
(seiscon 2012), 27-29 December , 2012.
[11] Bharti Wadhwa Amandeep Verma , Energy Saving Approaches for Green Cloud Computing: A Review,
978-1-4799-2291-8/14/$31.00 ©2014 IEEE
[12] Sonika P Reddy ,Chandan H K S, Energy Aware Scheduling of Real-Time and Non Real-Time Tasks on Cloud Processors
(Green Cloud Computing), ©2014 IEEE
[13] Zhuo Tang, Zhenzhen Cheng, Kenli Liztang@hnu.edu.cn Keqin LiAn ,Efficient Energy Scheduling Algorithm for Workflow Tasks in Hybrids and DVFS-enabled Cloud Environment,
© 2014 IEEE
[14] Er.Shimpy, Mr Jaydeep sindhu Different scheduling algorithms in different cloud computing, IJARCCE vol.3 issue 9 September 2013