Advance anticipatory performance improvement model for cloud computing

ADVANCE ANTICIPATORY PERFORMANCE IMPROVEMENT MODEL, FOR CLOUD COMPUTING Umesh Lilhore1,Dr Santosh Kumar2 Scholar MUIT Lucknow, Department of Computer Science & Engg.

2PhD Supervisor MUIT Lucknow, Department of Computer Science & Engg.

1PhD

Abstract: In cloud computing resources are presented as virtual machines. In such a scenario, scheduling algorithms play an important function where the intention is to schedule the tasks effectively and balance the entire load effectively, and can reduce the total response and execution time of the entire cloud system and improve the resource utilizations. In this research paper we presenting an “Advance Anticipatory Performance Improvement Model for Cloud Computing (AAPIMC)”, for cloud computing. Proposed “AAP-IMC” model uses combine strategy of our previous proposed methods “MFL-APSOM” model and “ADRS-ADDRM”, to optimize the total execution time of tasks in the workflow applications and load balancing. Proposed method compares with existing methods on various parameters and results clearly shows that proposed method have better performance results for cloud computing.

Keywords- Cloud Computing, Performance of Cloud computing, ADRS-DDMS, AAP-IMC, MFLAPSOM I. INTRODUCTION Distributed computing leads to a new technology called Cloud computing used by both academia and industry to store and retrieve the files and necessary documents. The main issue is to schedule the incoming requests in an efficient way with minimum response time and at the same time resources should not underutilized. Many algorithms like FCFS, Round Robin, Active-VM monitoring and Throttled are used for executing clients request with a minimum response time and also assigning the requests to the virtual machines [1]. But the constraints such as high communication delays, underutilization of the resources are not addressed clearly and efficiently, which leads to many of the resources does not participate in executing the requests and hence leads to imbalance of cloud system. The new era of cloud computing technology will focus on how effectively and efficiently the infrastructures are instantiated and available resources are utilized dynamically. In the Cloud, computing resources need to be allocated and scheduled their load in a way that providers higher and efficient utilization of computing resource and cloud users can meet their requirements for their applications performance with minimum expenditure. In this paper, we present an Anticipatory Performance Improvement Model for Cloud Computing which allocates incoming jobs to available virtual machines. Here the virtual machine assigned depending on its load i.e. VM with least request is found and then new request is allotted.

II. BACKGROUND AND RELATED WORK In this section, we briefly summarize the load balancing algorithms used in the cloud computing environment. The main focus is on the efficient utilization of the virtual machines and balancing the virtual machines with the incoming request. Load balancing is defined as a process of making effective resource utilization by reassigning the total load to the individual nodes of the collective system and thereby minimizing under or over utilization of the available resources or virtual machines. Hemant S. Mahalle, Parag R. Kaveri and Vinay Chavan [3] have developed Active monitoring load balancer algorithm which maintains information about each VMs and the number of requests currently allocated to which VM. When a request to

@IJRTER-2016, All Rights Reserved

210

International Journal of Recent Trends in Engineering & Research (IJRTER) Volume 02, Issue 08; August - 2016 [ISSN: 2455-1457]

allocate a new VM arrives, it identifies the least loaded VM. If there are more than one, the first identified is selected. Active VM Load Balancer returns the VM id to the Data Center Controller the data Center Controller sends the request to the VM identified by that id. Data Center Controller notifies the Active VM Load Balancer of the new allocation.

III. PROPOSED “AAP-IMC” MODEL In cloud computing better performance of the system is always desirable and challenging.

Performance of the cloud system depends on various factors such as resource availability, resource utilization and distribution, task scheduling and load balancing. Availability plays an important role in Cloud-based systems. Proposed performance improvement model for Cloud Computing (AAPIMC), presents a set of solutions for load balancing in the Cloud. Proposed model uses combine strategy of two proposed methods “MFL-APSOM” and “ADRS-ADDRM”, in various phases. To optimize the total execution time of tasks and instructions in the various workflow applications proposed method use following phases-

Figure 3 Proposed AAP-IMC Model, for Cloud

3.1 Performance parameter in proposed “AAP-IMC”- For performance comparison in between proposed and existing method following parameters is used. Each parameter are calculated secretly for both methods (proposed and existing), performances metrics and result comparisons for the cloud scheduling algorithms are based on following factors-

@IJRTER-2016, All Rights Reserved

211

International Journal of Recent Trends in Engineering & Research (IJRTER) Volume 02, Issue 08; August - 2016 [ISSN: 2455-1457]

Figure 3.1 Performance, parameter in proposed Model

3.2 Working steps of proposed AAP-IMC - Proposed method use following steps 1. Step 1-The load balancer gets the data from CSP. The load balancer is a middleware intermediate to CSP and data center.

2. Step 2-The middleware initially sends the request to data center to identify the failure and reliable VM within the time limit 1 to 1.5 ms. In data center there are multiple VMs are available. The data center receives the request form middleware and transfer s the request to all VMs in the i r data center. All the VMs respond to the middleware depending upon the status such as total memory, used memory, free memory, latency rate and execution rate are calculated 3. Step 3-The middleware lists the VM based on the response time except failure VM. The middleware allocates the load based on the priority. The priority is assigned 3.1. If the VMs respond within the time, the fewer time limits will take first and so on.

3.2. If two or more VMs respond at the same time limit, the less memory utilization VM needs as the next priority.

3.3. If the VM exceeds the time limit will assign as last priority and is the two or more VMs respond exceed the time limit the priority is assigned based on the IP address.

4. Step-4 Resource level percentage is calculated percentage. This process is done for each VM separately.

5. Step 5- Using the modified fuzzy rules the resource level is identified as low, medium and high.

6. Step 6- The load is assigned to the VM based on by with the percentage identified by the resource level percentage parameter.

IV. EXPERIMENTAL RESULTS FOR PROPOSED “AAP-IMC� MODEL The proposed algorithm uses 200 VMs in a single data centre. The memory for each VM can share the physical resources on a server. The different type of file size is given and the files allocated based on the resource level percentage taken in allocated VM. The proposed method is more efficient than the existing algorithm by increasing the number of VMs.

@IJRTER-2016, All Rights Reserved

212

International Journal of Recent Trends in Engineering & Research (IJRTER) Volume 02, Issue 08; August - 2016 [ISSN: 2455-1457]

4.1 Data Transmission Rate-How much amount of data is transfer in a particular time period. The table 4.1 and figure 4.1 shown below is the analysis of the total bits to be transmitted over a certain time of the communication, in between 100 VMs.

Table 4.1 Data Transmission Rate

Figure 4.1 Data Transmission Rate

4.2 Response Time- It is the amount of time taken from when a process is submitted until the first response is produced. Less response time shows better efficient performance.

Table 4.2 Response Time

@IJRTER-2016, All Rights Reserved

213

International Journal of Recent Trends in Engineering & Research (IJRTER) Volume 02, Issue 08; August - 2016 [ISSN: 2455-1457]

Figure 4.2 Response Time

4.3 Execution Time- The execution time is the total amount of time the virtual machine takes in executing the instructions. The time of an executing program, therefore, is generally much less than the total execution time of the program.

Table 4.3 Execution Time

Figure 4.3 Execution Time

4.4 Result Analysis- Results clearly shows that proposed, �AAP-IMC� performance method performs outstanding in terms of data transfer rate, execution time and response time.

V. CONCLUSION & FUTURE WORK The proposed work involves developing an efficient VM load balancing algorithm for the cloud and conducting a comparative analysis of the proposed algorithm with the existing algorithms.

@IJRTER-2016, All Rights Reserved

214

International Journal of Recent Trends in Engineering & Research (IJRTER) Volume 02, Issue 08; August - 2016 [ISSN: 2455-1457]

By studying the existing VM load balancing algorithms, the AAP-IMC proposes an efficient algorithm for VM load balancing.

REFERENCES Nuaimi, K.A. Al Ain. Mohamed, N. Nuaimi, M.A and Al Jaroodi,J.,“A Survey of Load Balancing in Cloud Computing: Challenges and Algorithms”, Published IEEE Network Cloud Computing and Applications (NCCA) Second Symposium , 2012.

2. Olivier Beaumont, Lionel Eyraud Dubois, Hubert Larchevque, “Reliable Service Allocation in Clouds”, IEEE 27th IEEE International Parallel & Distributed Processing Symposium (IPDPS) 2013.

3. P.Varalakshmi, Aravindh Rama swamy, Aswath Bala Subramanian and Palaniappan Vijay Kumar, “An Optimal Workflow Based Scheduling and Resource Allocation in Cloud”, Springer, pp 411- 420, year 2011.

4. Rohit O.Gupta and Tushar Champaneria, “A Survey of Proposed Job Scheduling Algorithms in Cloud Computing Environment”, International Journal of Advanced Research in Computer Science and Software Engineering, PP 782-790, year 2014.

5. S.Sindhu and Saswati Mukheree, “Efficient Task Scheduling Algorithms for Cloud Computing Environment”, ACM Transactions, Pages 79-83, July 2011.

6. Vilutis G., Daugirdas L., Kavaliunas R. and Sutiene K., “Model of load balancing and scheduling in Cloud computing”, IEEE Transactions, Pages 117-122, June 2012.

7. Wai-Leong Yeow, Cedric Westphal and las C. Kozat, “Designing and Embedding Reliable Virtual Infrastructures”, ACM, Pages 57-64, April 2011.

8. Xiaonian Wu, Mengqing Deng, Runlian Zhang, Bing Zeng and Shengyuan Zhoua, “A Task Scheduling Algorithm based on QoS Driven in Cloud Computing”, Science Direct, Pages 1162-1169, Jul 2013.

9. Ying Shao and Xiaoming Li,“Simulation and Analysis of the Reliability of the Cloud Model Based Automatic Environmental Monitoring System”, IEEE Power and Energy Engineering Conference (APPEEC), 2011 Asia Pacific Conference, Pages 21-27, March 2011.

10. Xu, Gaochao, Pang, Junjie and Fu, Xiaodong,” A load balancing model based on cloud partitioning for the public cloud”, IEEE Transactions, Pages-34-39, Feb 2013.

11. Yaojun Han and Xue mei Luo, “Hierarchical Scheduling Mechanisms for Multilingual Information Resources in Cloud Computing”, Springer, Pages 268-273, year 2013.

1.

@IJRTER-2016, All Rights Reserved

215