Fault Tolerant Cloud Systems

Fault Tolerant Cloud Systems

Sathish Kumar (VIT University, India) and Balamurugan B (VIT University, India)
Copyright: © 2018 |Pages: 16
DOI: 10.4018/978-1-5225-2255-3.ch093
OnDemand PDF Download:
List Price: $37.50


Cloud computing refers to a model for accessing computing resource like networks, servers, storage, applications and services by remotely. Cloud computing offers these resources as a service, namely infrastructure –as-a-service, platform-as-a-service, and software-as-a-service. To use these service two roles involved: the cloud provider offers the service and the cloud customer consumes the service. These resources are efficiently shared and utilized by customers and it is called workload. The requirement of workload depends on customer demands that vary from higher to lower. Based on the customer demand, cloud provider makes the resource available efficiently. In the context of cloud, the workload is based on web-based service or jobs processed in batch mode. The arrival process of jobs in the cloud is no often deterministic. The irregular increase or decrease in workload has a vital impact on resource provision. Monitoring the resources helps in measuring the performance of the cloud so that the resource can be provisioned to customers efficiently.
Chapter Preview


Computing is a study of algorithms, automation, programming the information. Programming is a way of designing algorithms which are aimed at controlling, executing the computing devices. These devices have the basic features such as the amount of data they can store and process speed to perform in a reliable time. Traditionally in 1980’s desktop personal computers (PCs) are used to support in creating, editing and manipulating documents. Further, these PCs are connected to the devices like a scanner to scan the documents, printer to take hard copies of the documents, etc. Later these devices are connected together to form a simple network. Since PCs has more of devices and it occupies more space the devices like laptop, tablet, mobile phone came into the context.

Figure 1.

Sample computing paradigm shift



Computing Shift from Mainframe to Cloud

There are five distinct stages that cloud computing arrived. Initially one computer terminals like keyboard monitor to access the mainframes systems. In stage1, personal computers (PCs) were used to manipulate user requirements. In stage2, several PCs were connected to form a network called local network and user can access the PCs from their own PCs. In stage3, several local networks were connected to a global network called the internet. From the internet, the users can remotely access the systems. In stage4, the grid computing came into the context were resources were shared distributedly. The user uses PCs to access the grid. In stage5, the user employs a computing technique called cloud computing that allows users to access the resources through the internet.

Figure 2.

Mainframe to cloud shift


Computing Techniques Era

Cluster Computing

A cluster computing consists of several stand-alone computers which are a distributed loosely or tightly connected system and performs several tasks which are viewed as a single system. The features of cluster computing are reducing cost, power; it uses improved network technology, availability, and scalability.

Service-Oriented Architecture (SOA)

Service-oriented architecture is a loosely coupled distributed system that follows standard protocols to provide services over a network. The aim of service-oriented architecture is to divide the problems into separate distinguishable sections and emphasizes in a single software. The feature of the service-oriented architecture is it is independent of any product or technology.

Grid Computing

Grid computing is the technology that came into the contrast of electrical power grid, which allows the resource to be shared independently between power grids. Grid computing is a loosely coupled distributed system connected over a network. Grid computers are geographically distributed and heterogeneous in nature. The feature of grid computing is it improves scalability and performance of information system.

Utility Computing

Utility computing is a technology makes computing resource available as a service and provides a resource on demand and charges them depends on the usage. Utility computing used some form of virtualization like storage virtualization and hardware virtualization. The feature of utility computing is in reducing the cost of buying a resource in rental.

Key Terms in this Chapter

Work Load: The number of resources consumed and shared by the users for a period of time.

Cloud Provider: Cloud providers are who provides the services to the customer who needs the cloud services.

Cloud Models: Cloud is basically deployed in three models as public, private and hybrid.

Cloud Services: Cloud provides their resources as services to their customers. So the customers pay for the services and use them.

Virtual Machine Migration: The process of migrating, the virtual machine from one location to another is called virtual machine migration. AU76: Reference appears to be out of alphabetical order. Please check

Cloud Customer: A cloud customer is who consumes the services provided by the cloud provider.

Complete Chapter List

Search this Book: