Distributed, Networked and Mobile Computing

Summary

The changing characteristics of data and need for new data processing models as well as file systems in high-performance computing environment have been discussed in previous chapters. The large data-sets produce not only structured data but un-structured data too. Storage requirement for unstructured data is entirely different from that of structured data. There is a need to maintain quicker data storage, search and retrieval as these are basic requirements in high-performance computing.

Cloud computing not only provides support for traditional DBMSs, modern data processing requirements are also catered to as well. This chapter focuses on the characteristics of this new type of databases and discusses how un-structured data are stored in those databases for efficient processing. Apart from these, the chapter also discusses about different forms of database solutions available on the high-performance cloud computing environment.

Data storage and database on the cloud is intimately tied with one another and that provides the scope for suitable solutions to optimize the database performance. This has changed the way how database is managed. Many cloud computing vendors have developed new methods of storing data objects which are different from the traditional methods of storing data.

Data storage and database on cloud like high-performance systems are often intimately tied with one another for efficient processing of large volume unstructured data-sets.

DATABASE IN CLOUD

Consumers can avail database facility in cloud in two forms. First one is the general database solution that is implemented through installation of some database solution on IaaS (virtual machine delivered as IaaS). The other one is delivered by service providers as database-as-a-service where the vendor fully manages the backend administration jobs like installation, security management and resource assignment tasks.

In the first approach, the users can deploy database applications on cloud virtual machines like any other applications software. Apart from this, the ready-made machine images supplied by the vendors are also available with per-installed and pre-configured databases. For example, Amazon provides ready-made EC2 machine image with pre-installed Oracle Database.

In the Database-as-a-Service (DBaaS) model, the operational burden of provisioning, configuration, backup facilities are managed by the service operators.

Summary

Resource virtualization technique creates room for the adoption of dynamic approach in computing resource provisioning. The dynamic resource provisioning approach in turn creates the scope for developing scalable computing systems and applications. Scalability of systems and applications is an essential feature of cloud computing.

Computing cost depends on the total volume of resources acquired by an application. Any acquired and unutilized resource unnecessarily increases computing cost. Again, low acquisition of resource may affect application performance during higher demand. Hence, any system must run with minimum volume of required resources and should have the ability to expand itself with growing workload which is critical from business point of view. Again, a system should also have the ability to reduce itself with declining workload in terms of acquired resources. Otherwise unnecessary resource acquisition increases the cost.

This ability of expanding and shrinking of a system as per workload is known as scaling. Dynamic resource provisioning plays a key role in building of a scalable system but that alone cannot ensure the scaling. A system should also have the ability to integrate the provisioned resources effectively into itself (or release extra resources) and still run as the same system without any interruption or hitch. This ensures smooth user experience and at the same time reduces cost of computing and improves performance of applications.

WHAT IS SCALING?

Scaling is the characteristic of a system, model or function that describes its ability of growing or shrinking whenever required. In computing, scaling represents the capability of a system or application to deal with varying workload efficiently without bringing in a situation where resource shortage hampers performance or resource surplus increases the computation cost.

In simple words, scaling is defined as the ability of being enlarged (or shrunk) for accommodating growth (or fall-off) to fulfill the business needs. A system or application architecture can be termed as scalable if its performance improves on adding new resources and the improvement is proportional to the capacity added.

A system that scales well can maintain its level of performance or efficiency when it works under larger or growing operational demands.

Summary

The emergence of cloud computing provides benefits of the utility service model to the computing users. Users of computing are now being called subscribers or consumers as they move towards cloud computing. Cloud computing is delivered to its subscribers over the internetwork as well as Internet. Subscribers can access the computing facility on subscription basis, anytime and anywhere.

The scores of benefits of cloud computing are attracting users towards it. But any new innovation comes with few challenges and cloud computing is not an exception. This chapter discusses various benefits of cloud computing and also presents the challenges before it.

The biggest challenge is related to data security and compliance issue. Most of the other critical challenges are due to the absence of open standards where vendors develop clouds using their own proprietary standard or technology. The good aspect is that, significant efforts have been undertaken to resolve all of these issues. Apart from these, this chapter briefly presents the role of web services in cloud computing development.

ORIGIN OF THE TERM ‘CLOUD COMPUTING’

The origin of the term ‘cloud computing’ dates back to the early 1990s. In those early days of network design, network engineers used to draw network diagrams representing different devices and connections among them. In such diagrams, they used to represent outer network arenas with cloud symbol since those details were not in their knowledge. This was known as ‘network cloud’ or ‘cloud’ in the networking industry during that period, but today we do not mean ‘cloud computing’ in the same sense.

With the beginning of utility computing initiatives towards the end of the last century, major software firms focussed on deliver applications over the Internet. Email services gained pace during this period as the vendors started to offer the facility to their users. And the most remarkable initiative came from Salesforce.com when they delivered business application for enterprises over the Internet in 1999. But all of these efforts were seen as part of utility computing facility development. Cloud computing did not emerge till then.

The term ‘cloud computing’ appeared in the market with its present meaning in the year 2006.

Summary

Cloud Computing has an alluring concept about service, and that is, it has the ability to deliver infinite resources: consumers can get any kind and any volume of resources instantly as per their demands. The somewhat incorrect idea of infinite resources in cloud computing has primarily been accomplished by creating flexible resource pools. Resource virtualization technique and auto-scaling mechanism enable an uninterrupted supply of resources during the execution of system or application.

However, if not understood properly the concept of infinite computing resources may cause serious concerns about the success of a cloud service. At the physical (data center) level, it is never possible for a cloud service provider to arrange an unlimited volume of computing resources. Service providers actually create the impression of unlimited resources before their consumers by strategic arrangement and utilization of resources. This chapter will discuss these strategies.

Earlier, the safest traditional approach in capacity planning was to buy resources for an estimated maximum capacity, which resulted in resource wastage and an unprecedented increase in budget. However with the assurance of an unlimited and dynamic supply of resources, enterprises can now plan business with minimum required resources. This reduces resource wastage as well as computing costs.

Apparently, it is the responsibility of the IaaS providers to deliver all resources as per the demand of the consumers. However without the sincere participation of the upper layer service providers (PaaS and SaaS providers) IaaS providers alone cannot make this idea successful. Even end users of cloud services (application consumers) have major roles to play. The chapter focuses on these aspects also.

WHAT IS CAPACITY PLANNING

Capacity Planning in computing is basically developing a strategy which guarantees that at any moment, the available or arranged resources will be sufficient to support the actual demand for resources and that too at the minimal possible cost. The goal of capacity planning is to identify the right amount of resource requirement to meet the service demands at present and also in the future.

Resource requirement of an application generally differs with time. A cost effective agile system can only be developed by understanding these shifting resource needs, and with proper capacity planning in place. Appropriate capacity planning made for a system offers enormous benefits.

Summary

The introduction of cloud computing has revolutionized the way of computing as it was being done. The impact of cloud computing technology has reached every corner of the world. It has introduced a new era where innovative technological advancements are emerging based on the strength and utility of cloud computing. Cloud computing may take some time to attain its full maturity but it has appeared as the biggest thing after the Internet. With its fast adoption in different domains and businesses, researchers are working on several issues of cloud computing. This concluding chapter of the book discusses such topics.

This chapter briefly introduces two topics which are emerging as subjects of attraction after cloud computing. Mobile cloud computing is a technology that has opened the scope for overcoming the limitations of mobile computing through effective utilization of cloud computing. This makes mobile computing attractive and more useful for users, especially for billions of smart phone users.

The most remarkable happening in the connectivity domain, after cloud computing is the ‘internet of things’. It promises to bring a revolution in human civilization by connecting living and non-living objects, and simplifying the act of collecting different information about them. All these digital innovations are propelling us towards a smarter planet.

MOBILE CLOUD COMPUTING

Mobile cloud computing (MCC) refers to the computing facility introduced by combining three fascinating technologies together: mobile computing, cloud computing and wireless communication. The emergence of MCC is followed by the explosive growth in uses of mobile devices, especially smart phones where cloud computing acts as a potential technology enabler for mobile services. The actual objective of MCC is to provide rich experience to users by offering cutting edge applications over a variety of mobile devices.

The popularity of mobile devices which facilitates trouble-free communication and instant availability of information has made them an integral part of modern age human civilization. As a consequence, mobile cloud computing has emerged as an advanced step in the process of enriching mobile computing.

Mobile Computing

In simple words, mobile computing is a technology where computing-capable devices can transmit data without being connected to any device or network physically, even when in transit.

Summary

Cloud computing does not imply any single technology. In short, a combination of multiple methodologies defines its architecture. The management of various aspects associated with this computing facility thus becomes very important case for successful outcome during its real-life implementations. Be it business delivery, or technical foundation of system development, the ‘service’ takes center stage in cloud computing. Consequently, cloud service management has become another important issue to be discussed in a new context.

Most management activities in cloud are automated and SLA-driven. The service level agreement (SLA) thus plays an important role in the success of cloud computing. This chapter represents the lifecycle of SLA, apart from discussing the lifecycle of cloud service. Cloud management standards, tools and responsibilities of providers and consumers have also been discussed.

This chapter also focuses on various programming models which are implemented on cloud. For this purpose a case study is represented with Aneka cloud platform which has a special feature of supporting multiple programming models. In this context, the chapter tries to revisit the cloud computing architecture once more and represents its design characteristics and few non-functional features as well from different angles.

CLOUD ARCHITECTURE: REVISITED

In very broad sense, the cloud computing system can be viewed as composition of two elements:

Summary

Management of any large system turns out to be a complex task if the architecture of the system is not flexible enough. It becomes difficult to deploy new functionalities, alter existing functionalities and integrate interoperability when an immovable system grows. As a solution, any large system should be decomposed into functional primitives for efficient and flexible management and this philosophy perfectly applies to cloud computing systems too.

Clouds are generally heterogeneous and large in volume by their own nature. Hence, the flexible architecture (or flexible application architecture) is an essential criterion for making the computing system agile and capable of adjusting to quick changes in business strategy to take advantage in a competitive business market.

The conventional application architectures are not designed to take advantage of infrastructural agility and diversity like a cloud. In cloud computing, applications perform best when they are built based on the paradigm called as ‘service-oriented architecture’ or SOA. This new architectural paradigm makes cloud applications flexible and specially shows its ability when the system grows over time. This chapter focuses on the philosophy of the service orientation, discusses its advantages and indicates how business application becomes agile with SOA implementation.

THE PRE-SOA ERA

Conventional computing systems were based on centrally-administrated architecture. But, such architecture was not appropriate for quick application development, neither was it designed to support rapid changes in functionality. Moreover, the processes of developing a new application system used to become complex, time-consuming and expensive.

Then the concept of component based application development (CBD) model emerged. It had many goals similar to SOA. A component is a software object which is meant to interact with other components. Components are used to represent basic system functionalities, those which collaborate with other components. CBD promotes decomposition and reusability. It also increases productivity, quality and decreases the time-to-market. But interoperability is the issue that could not be handled with this system development model.

With the widespread adoption of heterogeneous distributed systems, the need for application interoperability appeared. Heterogeneous system refers to the ability of integrating multiple type (cross architecture and/or cross vendor) of one or more computing resources (like processor) into a single computing environment. Cloud computing system is the ultimate outcome of heterogeneity.

Summary

Cloud computing is not an abrupt innovation. Rather, it is a series of developments that have taken place over past few decades. Progresses in computing technology, starting from its early days, has slowly metamorphosed into cloud computing in this advanced era. Although the idea of cloud computing originated long ago, the concept could not materialize due to lack of necessary technological elements.

Documentary evidence can be traced back to the 1960s, when John McCarthy (who coined the term ‘artificial intelligence’) wrote that ‘computation may someday be organized as a public utility.’ Since then, computing technology has gone through phases of development. Hardware and communication technology have been progressed, Internet has changed the world and at the same time, the web-based software architecture has also matured.

As advancements in all associated fields have slowly overcome the limitations of earlier approaches, it has been possible to realize the dream of computing as the new measure of public utility. This chapter focusses on the evolution of cloud computing and discusses how generations have developed through stages like centralized computing, client server computing, distributed computing and grid computing on to cloud computing.

THE EVOLUTION OF CLOUD COMPUTING

Cloud computing is not an isolated development. Cloud technology has been matured over the years with constant advancements in the field of computing. The beginning can be traced back to a time when remote access to time-shared computing system became a reality. The realization of cloud computing has been closely linked with several other subsequent developments in the domain.

Several decades of research, particularly in the domain of parallel and distributed computing, have paved the way for cloud computing.

A thorough discussion about the development of cloud computing can never overlook the continuous innovations in the field of electronic and computing hardware. As the hardware technology evolved, so did the software. Beside these, with the advancements in communication protocols, network communication technology as well as Internet also played a vital role in this process. This section focusses on different phases of developments in the field of computing starting from the mainframe age and discusses how those progresses have contributed towards the growth of cloud computing.

Summary

Cloud computing is about using services provided by cloud vendors. When consumers invest effort and capital to develop their computing setup on some provider's service, one major concern for them is whether they can move to other cloud service in future, if necessary, with minimal disruption and effort. Can things once built over one cloud service be moved to another cloud service? This question arises since there are possibilities of vendor-lock-in, a situation where a consumer using one cloud service cannot easily move to a similar service delivered by a competitor company or service provider. All investments by a consumer may be in stake if computing setup created on one provider's service cannot be moved to other provider's service in future.

An other concern comes in mind when consumers think about linking applications of two different cloud services together or, when it is required to link some on-premises non-cloud or cloud computing setup with a public cloud service. Such linking brings forth the issue of interoperability. Can applications of two different cloud environments be linked in such a way? This concern may worry the consumers.

Consumers should have a clear understanding regarding these issues before moving into cloud. Such understanding enables one to take informed decisions when choices have to be made concerning technology, application or platform.

Cloud adoption among the consumers also depends on how a cloud environment can address users’ concerns regarding portability and interoperability, apart from the issue of security.

CHALLENGES IN THE CLOUD

Although the cloud service offerings present a simplistic view of computing services to consumers, there are few critical issues to consider while moving into cloud. As already discussed, the prime challenges among these are the challenges associated with information security, privacy and compliance. The other vital challenges are related to issues of:

Summary

Security is one of the topmost concerns of any computing model and cloud computing is no exception. Consumers or enterprises moving into cloud need to feel secure about their computing facilities and more importantly about the data they are disclosing to service providers. This is a critical choice for enterprises accustomed to safeguarding data sitting in their own centers in the traditional way of computing. Cloud computing promotes the concept of working on proprietary data and applications outside their jurisdiction.

It is often said that, ‘cloud computing is an excellent idea, but cloud security is not good’. The common perception is that the cloud services are not inherently secure. Cloud computing creates scope for scores of possibilities although like any other technology, there are some risks associated with it which can be overcome if understood correctly.

A cloud computing environment can be formed in different ways and there can be several ‘cloud formations’. Each form is not suitable for every kind of business operations. Consumers must first understand which among these forms which is best-suited for their purpose. This little judgment can make the security of cloud computing even better than that of traditional in-house computing environments. This chapter focusses on these important aspects.

If studied and designed appropriately, cloud computing causes no more threat to security than what exist in traditional computing.

THE SECURITY CONCERN IN CLOUD

Traditional computing systems used to create a security boundary by placing firewalls at the gateways through which the network used to communicate with the outer world. Firewall blocks unwanted traffic trying to access the network it protects and thus only authenticated accesses are allowed into the system. Thus, malicious accesses get blocked at firewalls in the traditional data centers in order to keep the system protected from outside threats.

But this strategy only makes sense when all applications and data reside within one network of security perimeter. Traditional data centers allow perimeterized (i.e. within organization's own network boundary or perimeter) access to computing resources. But, the de-perimeterization (to open-up the interaction with outer network) and erosion of trust boundaries that was happening in enterprise applications, have been amplified and accelerated by cloud computing.

Summary

With appropriate capacity planning in cloud computing, the large pool of virtualized resources creates an illusion of an infinite source of computing resources. In the preceding chapters, it has been discussed how applications often need to scale to manage the varying workload and how resources are dynamically provisioned in order to meet the needs of scaling in cloud computing. It has also been discussed that horizontal scaling is more prepared for large scale scaling in distributed environment like cloud.

Upon obtaining provisioned or de-provisioned dynamic resources, the horizontal scaling provides elasticity to cloud computing. But, this flexibility comes with lots of implementation complexity. Load balancing is the mechanism through which these features are implemented. Apart from these, as there are multiple resources available to serve a particular type of service request, and in a distributed computing environment it becomes necessary to distribute proportionately those service requests among the available resources so that none of them becomes overloaded and degrades the performance of the entire system. It is the run-time logic or algorithm of distributing the service requests that has to be capable of evenly distributing the workload. This chapter discusses different types of load balancing techniques along with their pros and cons. The possible parameters for consideration while working out load balancing algorithms have also been discussed. A properly-designed workload distribution and balancing architecture reduces resource underutilization, improves the performance of a system and also brings down the cost of computation.

Load balancing is an important ingredient for building any dynamic, flexible, scalable and robustly-distributed computing architecture.

LOAD BALANCING

Load balancing in distributed computing systems is an essential technique to distribute processing and communication activities evenly across the resources in the network so that no single computing resource gets overloaded. Among different exercises, load balancing is especially important for applications which often deal with a large and unpredictable numbers of service requests. Load balancing is a technique that distributes load evenly among multiple computing resources such as processor, memory etc.

Through load balancing, the incoming service requests are distributed among available computing resources. An efficient load balancing mechanism improves average resource utilization rate and therefore enhances the overall performance of a system.

Summary

The easy access to high-performance computing resources in cloud computing has not only made the process-intensive activities smarter but the data-intensive computing activities also have taken center stage. The nature of data have radically changed with this revolutionary utility service; hence their processing and storage requirements vary. Large data-sets are generated and produced everyday are sent for processing in the high-performance computing environments.

Like the traditional storages, users can store and access multimedia files of various formats like text, image, audio and video in cloud also, but the storage requirements have been altered for efficient processing of the large data-sets which are produced in cloud every hour. The traditional enterprise level files and data storage systems were not sufficient to satisfy all of the data-intensive and high-performance computing requirements.

Efficient file handling to support parallel and distributed operations of large data-sets needed an entirely new file system format. Hence, researchers and computing vendors have come up with suitable storage solutions to achieve optimal performance in cloud like high-performance environment. This chapter focuses on all of these advancements made to fulfill those requirements.

Cloud computing promises high-performance. Hence, the file system and storage to support high-performance data processing are critical requirements of cloud environment.

REQUIREMENTS OF DATA-INTENSIVE COMPUTING

Data-intensive computing presents a challenge to computing systems in terms of delivering high-performance. Large volume complex data-sets cannot be processed centrally in a single node and require partitioning and distribution over multiple processing nodes. Thus, data-intensive computing is I/O-bound and requires rapid movements of data in large numbers. This requires appropriate management of data in transaction. Data modelling, partitioning, node assignment and accumulation are some of the critical parts of this computing. Consumers’ requirements and technical aspects related to the storage facility in high-performance computing environment are different from the traditional storage system in many ways. Traditional enterprise storage systems are no more sufficient to tackle those issues.

Scalability and high-performance distributed data processing are complementary to each other. Distribution of large data-set among as-many-nodes as required for processing promotes scalability of the application. Suitable file systems are required to support this distribution and scaling proficiently. It has been observed that data-intensive computing often involves process-intensive computing too. Complex data-sets present challenges before the computing system.

Summary

The future often arrives silently. In the digital world, cloud computing which emerged silently, has already made a significant place of its own and has shown enough potential to be the future of computing. The advantages it offers are plenty, and this is what makes cloud computing revolutionary.

Books on this evolving subject started appearing in the market towards the end of the last decade. A handful of books have been published so far, and they provide useful resource material. While a few among them have been written in order to discuss conceptual aspects of the subject, others have been written from research or application perspective. As the research or application oriented books serve specific purposes only, they are not appropriate for general readers of the subject.

Cloud computing is not a single technology; rather, it is a combination of multiple technologies. Hence, it is necessary to understand all those individual technologies in order to understand cloud computing as a whole. Many books have been written with their focus on some particular aspects of cloud computing, and thus, they have not been able to cover all the essential topics of cloud computing with equal importance. This book is written from conceptual perspective, for the sake of new learners of the subject, and it tries to fill the void for a book covering all the fundamentals in detail, with simplified explanations in a learner friendly manner.

The author felt the need for such a book while teaching the subject to the undergraduate engineering students, as well as software professionals who wanted to develop a basic understanding of the subject. It is either inadequate coverage or a complex approach of most books available in the market, that makes it difficult for new learners to develop a clear conception about the various aspects of these technologies, and they therefore do not satisfy the fundamental queries of students. Thus, it is a lack of books covering all the fundamentals in detail which encouraged the author to write this book.

The author is aware of the fact that computing and IT industry are fast merging into a singular cloud computing environment, and in the near future there will be a requirement for a large pool of workforce with sound insight into this subject.

Summary

Cloud computing delivers virtualized resources. In previous chapter, it has been discussed how virtualization technology has revolutionized the computing environment. It is important to note that in reality the virtualized resources are effective representations of actual physical resources. Cloud data centers maintain a large pool of physical computing resources which is presented in virtual mode as infrastructure service. This chapter focuses on how those resource pools are organized in cloud data centers.

Resource sharing is another important feature of cloud computing. Any resource component from the pool of resources is not dedicatedly used to serve one particular user or application; rather they are shared among multiple users or applications. Through this increment in resource utilization rate, a comparatively smaller number of resources can serve a large number of users.

Resource pooling, virtualization and sharing of resources ultimately converge to produce the dynamic behaviour of cloud systems. Cloud supplies the resources dynamically to users. This dynamic resource provisioning eliminates the need of permanent procurement of computing resources. This turns out to be a very effective technique as application demands or loads during computation do not remain same over a period of time.

RESOURCE POOLING

The utility service model of cloud computing requires to maintain huge amount of all types of computing resources to provide different services to consumers. For this purpose, cloud service providers create pool of computing resources. Effective pooling or grouping of resources requires appropriate system designing and architectural planning.

Resource pooling in cloud needs setting up of strategies by provider for categorizing and managing resources. Before resource pooling concept appeared, earlier people used to maintain discrete and independent set of resources known as silos.

In traditional computing model, the silos are made with very little or no inter-connections. On the other hand in cloud computing, the consumers use well-connected pool of computing resources. They gain almost no knowledge or control over the locations from where physical resources are allotted to them. In its out-of-best scenario, the providers sometimes ask for choice of geographic location (country or continent) from where a consumer wants to get resources. But this choice is only possible in case of large service providers who have data centers at multiple geographic locations around the world.

Distributed, Networked and Mobile Computing

Refine search

Refine search

Actions for selected content:

1983 results in Distributed, Networked and Mobile Computing

Frontmatter

14 - Database Technology

Summary

Tables

9 - Scaling in the Cloud

Summary

3 - Benefits and Challenges

Summary

10 - Capacity Planning

Summary

21 - Mobile Cloud Computing and The Internet of Things

Summary

19 - Cloud Management and a Programming Model Case Study

Summary

12 - Service-Oriented Architecture

Summary

2 - Evolution and Enabling Technologies

Summary

Bibliography

18 - Portability and Interoperability Issues

Summary

6 - Security Reference Model

Summary

11 - Load Balancing

Summary

Endnote

13 - File System and Storage

Summary

Preface

Summary

7 - Resource Virtualization

Index

8 - Resource Pooling, Sharing and Provisioning

Summary

Distributed, Networked and Mobile Computing

Refine search

Refine search

Actions for selected content:

Save Search

1983 results in Distributed, Networked and Mobile Computing

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary

Summary