Thursday, 29 May 2014



In a large Infrastructure-as-a-Service (IaaS) cloud, component failures are quite common. Such failures may lead to occasional system downtime and eventual violation of Service Level Agreements (SLAs) on the cloud service availability. The availability analysis of the underlying infrastructure is useful to the service provider to design a system capable of providing a defined SLA, as well as to evaluate the capabilities of an existing one. This paper presents a scalable, stochastic model-driven approach to quantify the availability of a large-scale IaaS cloud, where failures are typically dealt with through migration of physical machines among three pools: hot (running), warm (turned on, but not ready), and cold (turned off). Since monolithic models do not scale for large systems, we use an interacting Markov chain based approach to demonstrate the reduction in the complexity of analysis and the solution time. The three pools are modeled by interacting sub-models. Dependencies among them are resolved using fixed-point iteration, for which existence of a solution is proved. The analytic-numeric solutions obtained from the proposed approach and from the monolithic model are compared. We show that the errors introduced by interacting sub-models are insignificant and that our approach can handle very large size IaaS clouds. The simulative solution is also considered for the proposed model, and solution times of the methods are compared.
Cloud availability analysis can be performed through proper modeling techniques. Among these, state-space models are popular as they capture complicated interactions among system components and different failure/repair behaviors. However, for a large IaaS cloud, the model state space tends to be too large. Monolithic or one-level Markov chains are a typical modeling formalism, representative of the state-of-the-art in cloud availability modeling. The growth of the state space as the model takes into account more details of the system is known as the largeness problem of Markov models. Many of the existing published models, are hierarchical in nature.
v Stochastic Petri Nets (SPNs) can be used to tolerate the largeness problem, as they allow the automated generation of the (underlying) Markov model. Still, the solution of large models is an issue.
v Complexity and characteristics of large IaaS clouds (e.g., migration of PMs from one pool to another) lead to cyclic dependency among the sub models, needing fixed-point iteration.
v We state the availability assessment problem for an IaaS cloud and propose a
In this paper, we consider an IaaS cloud where physical machines (PMs) are grouped into three pools based on power consumption and provisioning delay characteristics. PMs can be migrated from one pool to another due to failure/repair events. We evaluated the availability of a similar system in, where some simplistic assumptions restricted the applicability to limited scenarios. After relaxing those simplistic model assumptions, we
first follow the common approach of developing a single monolithic model using a variant of SPNs called Stochastic
Reward Nets (SRNs). However, such a monolithic model is not scalable for large clouds due to the large state space of the underlying Markov chain.
v To resolve large state space, we present a scalable stochastic modeling approach based on interacting sub-models.
v To resolve state-of-the-art  problem we propose realistic monolithic model
v The overall model solution is obtained by fixed-point iteration over individual sub-model solutions.




ü Processor                  -        Pentium –IV

ü Speed                        -        1.1 Ghz
ü RAM                         -        512 MB(min)
ü Hard Disk                 -        40 GB
ü Key Board                -        Standard Windows Keyboard
ü Mouse                       -        Two or Three Button Mouse
ü Monitor                     -        LCD/LED


         Operating system :         Windows XP.
         Coding Language :         C# .Net
         Data Base             :         SQL Server 2005
         Tool                     :         VISUAL STUDIO 2008.

Rahul Ghosh, Francesco Longo, Flavio Frattini, Stefano Russo, and Kishor S. Trivedi “SCALABLE ANALYTICS FOR IAAS CLOUD AVAILABILITY” IEEE TRANSACTIONS ON CLOUD COMPUTING, VOL. 2, NO. 1, JANUARY-MARCH 2014

No comments:

Post a Comment