07-09-2016, 11:17 AM
1453445342-myproject.docx (Size: 165.48 KB / Downloads: 5)
Abstract-cloud computing now a day is leading edge for the IT companies each and every business trying to minimize their operational cost as much as they can. There is lot of competition among the various cloud service providers. There are lot of aspects are there on the basis of that they enhance their service. Some of these aspects are ELASTICITY. This paper mainly deals with the various kind of elasticity in the cloud and how to improve the elasticity in the cloud so that the better service ,fast, reliable service can be obtained through it.
Introduction:
Cloud computing platforms are trying to gain a lot of attention toward it. Some innovative and new startup companies extensively using it for excellent growing computational needs. Various traditional IT companies adopting the cloud services for their growth in their infrastructure. Cloud providing the magnetic platforms for their users; for example, cloud platform available at low cost which is quite affordable by the organization (cheap electricity, less rates of administration labor, etc) and this is achievable at high availability by extreme distribution and replication. One dominant feature that is selling point for cloud vendors is the claim that they are elastic; the user has to only pay for the amount of service they use. In particular startup companies are also very trendy toward the service of cloud computing because at reasonable cost they get what they are applying to their business. Scalability is another term which is generally used with elasticity. We can be very sure that no one are accurate to let a customer pay for exact what they use and need. All providers apply a minimal charge .adapting the platform to sudden increase in workload may be delayed and so on. Thus we would want to know how elasticity they would provide to each system (i.e. user). As yet, the exact measurement of elasticity cannot be done and how the elasticity can be improved. This is the narrow gap that my paper addresses to you. We draw your attention to the difference between elasticity and scalability. Scalability is the system capability, network, or process to handle how the amount of work growing, and its potential to be flexible in order to control that growth. All the cloud providers highlight the elasticity they provide this is part of where elasticity needs, but perfect scalability does not confirm perfect elasticity. Time is a central thing in elasticity, which depends on the how responsive to change load of work, while scalability helps the system with the load up to what extent it would increase. good elasticity have other aspects also on the basis of which it is judged that is system cost reduced when workload is under scale or not utilized, while scalability just give thought to the growth.
ELASTICITY
Lets understand what is elasticity and how elasticity in cloud computing make it efficient in its service. As in any advertising likely field, it is important to highlight the marketable prone terms such as “elastic”. Several determined attempt have trying to explain the meaning of the term elastic, and other concern about how it used about cloud computing platforms, and along with brief discussion, they point to conclusion that the performance of the platform which is major for elasticity. Elasticity means ability to regain it original property after its use according to National Institute of Standards and Technology (NIST) elasticity is the raid Deprovisioning and provisioning of the capability having infinite resource capacity with no limitation of purchasing amount of server either it is software, infrastructure or platform. Or we can define It the granularity of the usability accentuation. Some of the researchers have trying to explain the elasticity numerically which will clarify the meaning how elasticity is measured. To measure the imperfect cost elasticity when running in a given workload, with different penalties for over and under provisioning; the sum of these is the penalty measurement for the workload. Joe weinman is one of the researchers who tried it.
Weinman starts from a simple theoretical base model: computational capacity is a resource assumed, then function of time indicates the demand curve, how much of the artifice is needed for working of the allocation rarely. Mathematical function treated it Demand (t). Similarly, a function Resource (t) shows each time how much of the resource is designated to the application. Perfect elasticity mean Resource(t) = Demand(t) at each value of time i.e. t. a situation is identified by the Weismann where resource(t >demand(t)“.here the over provisioning or extra resource come into the market and corresponding to it extra cost is applied over to the consumer. Resource gets wasted underutilized. Likely one another case where Demand (t) < Resource (t), and linear in the difference measures this by opportunity rice which, although the constant of the proportionality of the resource is much higher for underutilized demand than for over resource. Let’s demonstrates it through a figure in each CPU has a single resource it shows a hypothetical situation in which the sine wave indicates the demand of resource where as linear line shows the supply which is keep on increasing