PROBABILISTIC CONSOLIDATION OF VIRTUAL MACHINESIN SELF-ORGANIZING CLOUD DATA CENTERS

ABSTRACT:

Power efficiency is one of the main issues that will drive the design of data centers, especially of those devoted to provideCloud computing services. In virtualized data centers, consolidation of Virtual Machines (VMs) on the minimum number of physicalservers has been recognized as a very efficient approach, as this allows unloaded servers to be switched off or used to accommodatemore load, which is clearly a cheaper alternative to buy more resources. The consolidation problem must be solved on multipledimensions, since in modern data centers CPU is not the only critical resource: depending on the characteristics of the workload otherresources, for example, RAM and bandwidth, can become the bottleneck. The problem is so complex that centralized and deterministicsolutions are practically useless in large data centers with hundreds or thousands of servers. This paper presents ecoCloud, a selforganizingand adaptive approach for the consolidation of VMs on two resources, namely CPU and RAM. Decisions on the assignmentand migration of VMs are driven by probabilistic processes and are based exclusively on local information, which makes the approachvery simple to implement. Both a fluid-like mathematical model and experiments on a real data center show that the approach rapidlyconsolidates the workload, and CPU-bound and RAM-bound VMs are balanced, so that both resources are exploited efficiently.

EXISTING SYSTEM:

In the past few years important results have beenachieved in terms of energy consumption reduction,especially by improving the efficiency of cooling and powersupplying facilities in data centers. The Power UsageEffectiveness (PUE) index, defined as the ratio of theoverall power entering the data center and the powerdevoted to computing facilities, had typical values between2 and 3 only a few years ago, while now big Cloudcompanies have reached values lower than 1.1. However,much space remains for the optimization of the computingfacilities themselves. It has been estimated that most of thetime servers operate at 10-50 percent of their full capacity[2], [3]. This low utilization is also caused by the intrinsicvariability of VMs’ workload: the data center is planned tosustain the peaks of load, while for long periods of time (forexample, during nights and weekends), the load is muchlower [4], [5]. Since an active but idle server consumesbetween 50 and 70 percent of the power consumed when itis fully utilized [6], a large amount of energy is used even atlow utilization.

DISADVANTAGES OF EXISTING SYSTEM:

  • Itis power consuming.
  • Large amount of energy is used even atlow utilization.

PROBLEM STATEMENT:

The ever increasingdemand for computing resources has led companiesand resource providers to build large warehouse-sized datacenters, which require a significant amount of power to beoperated and hence consume a lot of energy.

SCOPE:

The optimalassignment of VM’s to reduce the power consumption.

PROPOSED SYSTEM:

We presented ecoCloud, an approach for consolidatingVMs on a single computing resource, i.e., the CPU. Here, the approach is extended to the multidimensionproblem, and is presented for the specific case in whichVMs are consolidated with respect to two resources: CPUand RAM. With ecoCloud, VMs are consolidated using twotypes of probabilistic procedures, for the assignment and themigration of VMs. Both procedures aim at increasing theutilization of servers and consolidating the workloaddynamically, with the twofold objective of saving electricalcosts and respecting the Service Level Agreements stipulatedwith users. All this is done by demanding the keydecisions to single servers, while the data center manager isonly requested to properly combine such local decisions.The approach is partly inspired by the ant algorithms usedfirst by Deneubourg et al. [9], and subsequently by a wideresearch community, to model the behavior of ant coloniesand solve many complex distributed problems. The characteristicsinherited by such algorithms make ecoCloudnovel and different from other solutions. Among suchcharacteristics: 1) the use of the swarm intelligence paradigm,which allows a complex problem to be solved bycombining simple operations performed by many autonomousactors (the single servers in our case); 2) the use ofprobabilistic procedures, inspired by those that model theoperations of real ants; and 3) the self-organizing behavior of system, which ensures that the assignment of VMs to servers dynamically adapts to the varying workload.

ADVANTAGES OF PROPOSED SYSTEM:

  • Efficient CPU usage.
  • It reduces power consumption.
  • Efficient resource utilization.

SYSTEM ARCHITECTURE:

SYSTEM CONFIGURATION:-

HARDWARE REQUIREMENTS:-

Processor-Pentium –IV

Speed-1.1 Ghz

RAM-512 MB(min)

Hard Disk-40 GB

Key Board-Standard Windows Keyboard

Mouse-Two or Three Button Mouse

Monitor-LCD/LED

SOFTWARE REQUIREMENTS:

•Operating system:Windows XP

•Coding Language:Java

•Data Base:MySQL

•Tool:Net Beans IDE

REFERENCE:

Carlo Mastroianni, Michela Meo andGiuseppe Papuzzo“Probabilistic Consolidation of Virtual Machinesin Self-Organizing Cloud Data Centers”IEEE TRANSACTIONS ON CLOUD COMPUTING, VOL. 1, NO. 2, JULY-DECEMBER 2013.