See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/258442011

The Distributed Computing Paradigms: P2P, Grid, Cluster, Cloud, and Jungle

Article · November 2013

Source: arXiv

CITATIONS READS

37 4,142

2 authors:

Brijender Kahanwal Tejendr Pratap Singh

Pt. C. L. S. GOVT. P. G. COLLEGE, SEC - 14, KARNAL Council of Scientific and Industrial Research (CSIR), New Delhi
12 PUBLICATIONS 68 CITATIONS 213 PUBLICATIONS 3,273 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

NICRA, CRIDA, HYDERABAD View project

All content following this page was uploaded by Brijender Kahanwal on 24 August 2019.

The user has requested enhancement of the downloaded file.

 International Journal of Latest Research in Science and Technology ISSN (Online):2278-5299
Vol.1,Issue 2 :Page No.183-187 ,July-August(2012)
http://www.mnkjournals.com/ijlrst.htm

The Distributed Computing Paradigms: P2P, Grid,

Cluster, Cloud, and Jungle
Brijender Kahanwal* Tejinder Pal Singh
Assistant Professor, CSE Department, Galaxy Global Group Assistant Professor, Applied Sciences Department, RPIIT,
of Institutions, Dinarpur, Ambala, Haryana Bastara, Karnal, Haryana
(imkahanwal@gmail.com) (tps5675@gmail.com)
INDIA INDIA

Abstract: The distributed computing is done on many systems to solve a large scale problem. The growing of high-speed
broadband networks in developed and developing countries, the continual increase in computing power, and the rapid growth of the
Internet have changed the way. In it the society manages information and information services. Historically, the state of computing
has gone through a series of platform and environmental changes. Distributed computing holds great assurance for using computer
systems effectively. As a result, supercomputer sites and datacenters have changed from providing high performance floating point
computing capabilities to concurrently servicing huge number of requests from billions of users. The distributed computing system
uses multiple computers to solve large-scale problems over the Internet. It becomes data-intensive and network-centric. The
applications of distributed computing have become increasingly wide-spread.
In distributed computing, the main stress is on the large scale resource sharing and always goes for the best performance. In
this article, we have reviewed the work done in the area of distributed computing paradigms. The main stress is on the evolving area
of cloud computing.

Keywords – Distributed Computing Paradigms, cloud, cluster, grid, jungle, P2P.

1 Introduction transparency as Location, Migration, Replication,

The growing popularity of the Internet and the Concurrency, and Parallelism. Flexibility should be easy to
availability of powerful computers and high-speed networks develop. Reliability encompasses some factors like no data
as low-cost commodity components are changing the way we loss, secure system, and fault tolerant systems. Performance
should be high. Scalability should scale indefinitely.
do computing. Distributed computing has been an essential
component of scientific computing for decades. It consists of a Distributed
set of processes that cooperate to achieve a common specific Computing
goal. It is widely recognized that Information and
Peer-to-Peer Cluster Utility Jungle
Communication Technologies (ICTs) have revolutionized the Computing Computing Computing Computing
everyday practice. Social networks represent a stepping stone
in the on-going process of using the Internet to enable the Grid
social manipulation of information and culture. Mostly social Computing
network sites are implemented on the concept of large
distributed computing systems. These are running in centrally Cloud
Computing
controlled data centers. However, the trend in these massively
scalable systems is toward the use of peer-to-peer, utility, Figure 1.1: Classification of Distributed Computing
cluster, and jungle computing. The utility computing is
basically the grid computing and the cloud computing which This review paper covers the distributing technologies.
is the recent topic of research. This classification is well In the section 3rd peer-to-peer computing is elaborated; in
shown in the Figure 1.1. section 4th, the cluster computing; in section 5th utility
With the increasing heterogeneity of the underlying computing which has the subsections about grid computing
hardware, the efficient mapping of computational problems and the cloud computing; and in section 6th, the jungle
onto the ’bare metal’ has become vastly more complex. There computing. This paper gives a good introductory knowledge
are many challenges of distributed computing as follows: about the distributing computing.
Transparency means to hide distribution from the users
at the high levels and to hide the distribution from the 2 Related Works
programs at the low levels. There are more forms of

ISSN 2278-5299 183

 International Journal of Latest Research in Science and Technology ISSN (Online):2278-5299
Vol.1,Issue 2 :Page No.183-187 ,July-August(2012)
http://www.mnkjournals.com/ijlrst.htm

The computing industry is one of the fastest growing

industries and it is stimulated by the rapid technological
developments in the areas of computer hardware and software.
The technological advances in hardware include chip
development and fabrication technologies, fast and cheap
microprocessors, as well as high bandwidth and low latency
interconnection networks. Among them, the recent advances Figure 3.1: P2P Network
in electronics technology have played a major role in the
development of powerful sequential and parallel computers. 4 Cluster Computing
Software technology is also developing fast. Mature A cluster computing comprises a set of independent or
software, such as Operating Systems, programming stand-alone computers and a network interconnecting them. It
languages, development methodologies, and tools, are now
works cooperatively together as a single integrated computing
available. This has enabled the development and deployment
of applications catering to scientific, engineering, and resource. A cluster is local in that all of its component
commercial needs. It should also be noted that grand subsystems are supervised within a single administrative
challenging applications, such as weather forecasting and domain, usually residing in a single room and managed as a
earthquake analysis, have become the main driving force single computer system. The components of a cluster are
behind the development of powerful parallel computers. connected to each other through fast local area networks. To
Distributed systems can be considered conventional
handle heavy workload with large datasets, clustered
networks of independent computers. They have multiple
system images, as each node runs its own operating system, computer systems have demonstrated impressive results in the
and the individual machines in a distributed system could be, past. The architecture of the cluster computing environment is
for example, combinations of Massively Parallel Processors shown in the Figure 4.1.
(MPPs), Symmetric Multiprocessors (SMPs), clusters, and
individual computers. Parallel Applications

Cloud services are mainly divided into three services Sequential Applications Parallel Programming Environment
delivery models: SaaS (Software as a Service, e.g. Google
Mail), PaaS (Platform as a Service, e.g. Google AppEngine) Cluster Middleware
(Single System Image and Availability Infrastructure)
and IaaS (Infrastructure as a Service, e.g. Amazon EC2).
Since the work presented in this chapter is strongly related to PC workstation PC workstation PC workstation
Infrastructure as a Service model, we only focus in this Communication S/W Communication S/W Communication S/W
Network Interface H/W Network Interface H/W Network Interface H/W
section on this category of service. IaaS providers aim to offer
resources to users as pay-as-you-go manner. A key provider of
such a service is Amazon through its Elastic Cloud HIGH SPEED NETWORK OR SWITCH
Computing (EC2) and Simple Storage Service (S3).
The academicians and the giant groups are doing their Figure 4.1: Cluster Computer Architecture
best to comeback with the new concepts of the distributing
computing and they have given so many good results. But
there is always an intension to develop a better technology, so Components of Cluster Computing:
we are here and cloud computing is the recent topic on which There are so many components of the cluster
work is in progress. computing as follows:
o High Performance Computers like PCs, Workstations etc.
o Micro- kernel based operating systems.
3 Peer-to-Peer Computing
o High speed networks or switches like Gigabit Ethernets.
Peer-to-peer (P2P) networking has been working o NICs (Network Interface Cards)
primarily on the scalability issues inherent in distributing o Fast Communication Protocols and Services
resources over a large number of networked processes. In a o Cluster Middleware which is hardware, Operating system
P2P system, every node acts as both a client and a server, kernels, applications and subsystems.
providing part of the system resources. Peer machines are o Parallel Programming Environment Tools like compilers,
simply client computers connected to the Internet. All client parallel virtual machines etc.
machines act autonomously to join or leave the system freely. o Sequential and Parallel applications
This implies that no master-slave relationship exists among The cluster middleware is very much capable for
the peers. No central coordination or no central database is
offering an elusive and a unified system image.
needed. In other words, no peer machine has a global view of
the entire P2P system. The system is self-organizing with There is the classification of clusters as shown in the
distributed control as shown in the Figure 3.1. Figure 4.2.

ISSN 2278-5299 184

 International Journal of Latest Research in Science and Technology ISSN (Online):2278-5299
Vol.1,Issue 2 :Page No.183-187 ,July-August(2012)
http://www.mnkjournals.com/ijlrst.htm

of extra hardware that is required to implement load

Cluster
Computing balancing. Combining the two functions, however, also makes
the device more difficult to troubleshoot.

Highly Avalilability Load Balancing High Performance High Performance Computing Clusters
Clusters Clusters Computing Clusters The enterprises are now using large-scale clusters that
are often shared across departments with easy public access.
High performance clusters are used where time to solution is
important. They are also used in cases where a problem is so
Software-based big it can’t “fit” on one single computer. To increase
computing throughput, HPC clusters are used in a variety of
ways.
Hardware-based The easiest way is to allow the cluster to act as a
compute farm. Instead of running a job on a local workstation,
it is submitted to the cluster for execution. The cluster will
Figure 4.2: Cluster Computing Classification manage the resources needed for the job and assign the job to
a work queue. When the resource (for instance, a server) is
High-availability clusters
These are also called Failover Clusters. These are available, the job gets executed and the results are returned to
groups of computers that support server applications. These the user. Users who need to run many similar jobs with
can be reliably utilized with a minimum of down-time. They different parameters or data sets find clusters ideal for this
operate by harnessing redundant computers in groups or kind of work. They can submit hundreds of jobs and allow the
clusters that provide continued service when system cluster to manage the work flow. Depending on the resources,
components fail. Without clustering, if a server running a all the jobs may run at the same time or some may wait in the
particular application crashes, the application will be queue while other jobs finish. This type of computing is local
unavailable until the crashed server is fixed. Such type of to a cluster node, which means the node doesn’t communicate
clusters remedies this situation by detecting with other nodes, but may need high speed file system access.
hardware/software faults, and immediately restarting the
application on another system without requiring 5 Utility Computing
administrative intervention, a process known as failover. As Utility computing is envisioned to be the next
part of this process, clustering software may configure the generation of Information Technology evolution that depicts
node before starting the application on it. how computing needs of users can be fulfilled in the future IT
industry. Its analogy is derived from the real world where
Load balancing clusters service providers maintain and supply utility services, such as
Load balancing is often required when building electrical power, gas, and water to consumers. Consumers in
solutions that handle large volumes of client requests or that turn pay service providers based on their usage. Therefore, the
have high demands on security and redundancy. Clusters underlying design of utility computing is based on a service
support multiuser and multitasking environments. These provisioning model, where users (consumers) pay providers
factors accompanied by the heterogeneous nature of the for using computing power only when they need to. Utility
cluster hardware and software leads to the situation that the computing focuses on a business model, by which customers
overall distribution of the workload of a cluster is hard to receive computing resources from a paid service provider. All
predict at any particular moment. The static approach to I/O grid/cloud platforms are regarded as utility service providers.
planning is consequently almost useless. The two major However, cloud computing offers a broader concept than
categories of load balancing implementations are: utility computing.
Software-based load balancing consists of special
software that is installed on the servers in a load-balanced 5.1 Grid Computing
cluster. The software dispatches or accepts requests from the The aim of Grid computing is to enable coordinated
client to the servers, based on different algorithms. The resource sharing and problem solving in dynamic, multi-
algorithms can be a simple round-robin algorithm or a much institutional virtual organizations.
more complicated algorithm that considers server affinity. For As an electric-utility power grid, a computing grid
example, Microsoft Network Load Balancing is a load offers an infrastructure that couples computers,
balancing software for Web farms, and Microsoft Component software/middleware, special instruments, and people and
Load Balancing is a load balancing software for application sensors together. Grid is often constructed across LAN, WAN,
farms. or Internet backbone networks at regional, national, or global
Hardware-based load balancing consists of a scales. Enterprises or organizations present grids as integrated
specialized switch or router with software to give it load computing resources. They can be viewed also as virtual
balancing functionality. This solution integrates switching and platforms to support virtual organizations. The computers
load balancing into a single device, which reduces the amount used in a grid are primarily workstations, servers, clusters, and

ISSN 2278-5299 185

 International Journal of Latest Research in Science and Technology ISSN (Online):2278-5299
Vol.1,Issue 2 :Page No.183-187 ,July-August(2012)
http://www.mnkjournals.com/ijlrst.htm

supercomputers. Personal computers, laptops and PDAs can

be used as access devices to a grid system. The grids can be of Cloud
many types as; Knowledge, Data, Computational, Application Computing
Service Provisioning, Interaction or Utility.
These have many pros and cons. Pros are like; these Private Hybrid Community
are capable to solve larger, more complex problems in a Public Clouds
Clouds Clouds Clouds
shorter time, these are easier to collaborate with other
organizations, and these make better use of existing hardware. Figure 5.2: Classification of Cloud Computing
Cons are like; Grid software and standards are still evolving,
learning curve to get started, and non-interactive job Public clouds in this deployment the cloud
submission. infrastructure is accessible to general public and shared in a
pay as you go model of payment. The cloud resources are
5.2 Cloud Computing accessible via the internet and the provider is responsible for
Cloud computing is another form of utility computing. ensuring the economies of scale and the management of the
It is a new term in the computing world and it signals the shared infrastructure. In this model clients can choose security
advent of a new computing paradigm. This new paradigm is level they need, and negotiate for service levels. Amazon Web
quickly developing and attracts a number of customers and Services EC2 is a public cloud. It is accessible to the general
vendors alike. The quick development of cloud computing is public.
being fuelled by the emerging computing technologies which Private clouds are another deployment model for cloud
allows for reasonably priced use of computing infrastructures services. In this model the cloud resources are not shared by
and mass storage capabilities. It also removes the need for unknown third parties. The cloud resources in this model may
heavy upfront investment in Information Technology (IT) be located within the client organization premises or offsite. In
infrastructure. this model the client security and compliance requirements are
Cloud computing is a computing paradigm that not affected though this offering does not bring the benefits
involves outsourcing of computing resources with the associated with reduced capital expenditure in IT
capabilities of expendable resource scalability, on-demand infrastructure investments. In this type of cloud the general
provisioning with little or no up-front IT infrastructure public does not have access to the private cloud neither does
investment costs. Cloud computing offers its benefits through the organization use the public cloud.
three types of service or delivery models namely Hybrid clouds as its name implies is a model of
infrastructure-as-a-service (IaaS), platform-as-a-service deployment which combines different clouds for example the
(PaaS) and software-as-a-Service (SaaS) as shown in the private and public clouds. In this model the combined clouds
Figure 5.1. retains their identities but are bound together by standardized
technology. In this type of cloud the general public does not
have access to the cloud, but the organization uses
infrastructure in both the public and private cloud.
Application
Community clouds are the fourth deployment model
(S/W as a Service
that can be used to deliver cloud computing services. In this
(SaaS))
model the cloud infrastructure is shared by multiple
(Business Computing)
organizations or institutions that have a shared concern or
interest such as compliance considerations, security
requirements. This type of cloud may be managed by the
Development Environment organization or by a third party and may be located on-
(Plateform as a Service(PaaS)) premises or off-premises. In this type of cloud both the public
(Framework Computing) and the organizations forming the community cloud have
Manage Resources access to the cloud services offered by the community cloud.
(Infrastructure as a Service (IaaS),
Communication as a Service (CaaS), & 6 Jungle Computing
Data-storage as a Service (DaaS)) Jungle computing is a simultaneous combination of
(Utility Computing) heterogeneous, hierarchical, and distributed computing
Figure 5.1: Basic concept of cloud computing model and services resources. In many realistic scientific research areas, domain
experts are being forced into concurrent use of multiple
It also delivers its service through four deployment clusters, grids, clouds, desktop grids, independent computers,
models namely, public cloud, private cloud, community cloud and more. Jungle computing refers to the use of diverse,
and hybrid cloud as the classification is shown in the Figure distributed and highly non-uniform high performance
5.2. computer systems to achieve peak performance. These new
distributed computing paradigms have led to a diverse
collection of resources available to research scientists,

ISSN 2278-5299 186

 International Journal of Latest Research in Science and Technology ISSN (Online):2278-5299
Vol.1,Issue 2 :Page No.183-187 ,July-August(2012)
http://www.mnkjournals.com/ijlrst.htm

including stand-alone machines, cluster systems, grids, clouds, References

desktop grids, etc. as shown in the Figure 6.1 and this varied [1] B. Amedro, F. Baude, D. Caromel, C. Delbe, I. Filali, F. Huet, E.
collection is named as jungle computing. Mathias, and O. Smirnov, “Cloud Computing: Principles, Systems
and Applications”, series Computer Communications and Networks,
Springer, pp. 163-178, 2010.
CLUSTERS [2] Chee Shin Yeo, Rajkumar Buyya, Marcos Dias de Assunção, Jia Yu,
GRIDS Anthony Sulistio, Srikumar Venugopal, and Martin Placek, “Utility
CLOUDS Computing on Global Grids”, Chapter 143, Hossein Bidgoli (ed.),
The Handbook of Computer Networks, ISBN: 978-0-471-78461-6,
John Wiley & Sons, New York, USA, 2007.
[3] Cloud Security Alliance (CSA), “Security Guidance for Critical
Areas of Focus in Cloud Computing V2.1”, 2009.
INDEPENDENT JUNGLE COMPUTING DESKTOP [4] Douglas Eadline, “High Performance Computing For Dummies, Sun
COMPUTERS GRIDS and AMD Special Edition”, Wiley Publishing, Inc., Indianapolis,
Indiana, 2009.
PEER 2 PEER [5] F. J. Seinstra, J. Maassen et al., “Jungle Computing: Distributed
Supercomputing beyond Clusters, Grids, and Clouds”, Department of
Computer Science, Vrije Universiteit, De Boelelaan 1081A, 1081 HV
Network
Amsterdam, The Netherlands, pp. 1-31, 2010.
[6] Frank J. Seinstra, Jason Maassen, Rob V. van Nieuwpoort, Niels
Figure 6.1: The jungle computing - a diverse collection of computing Drost, Timo van Kessel, Ben van Werkhoven, Jacopo Urbani, Ceriel
Jacobs, Thilo Kielmann, and Henri E. Bal, “Jungle Computing:
The increasing complexity of the high performance Distributed Supercomputing beyond Clusters, Grids, and Clouds”, In
M. Cafaro and G. Aloisio, editors, Grids, Clouds and Virtualization,
computing environment has provided a bewildering range of Springer-Verlag '10, pp. 1-31. 2010.
choices beside traditional supercomputers and clusters. [7] H. E. Bal, J. Maassen, R. V. van Nieuwpoort, N. Drost, R. Kemp, N.
Palmer, T. Kielmann, F. Seinstra, and C. Jacobs, “Real-world
Scientists can now use grid and cloud infrastructures, in a distributed computing with Ibis”, Computer, Vol. 43, No. 8, pp. 54-
variety of combinations along with traditional supercomputers 62, 2010.
[8] Hwang, Dongarra, and Fox, “Distributed and Cloud Computing”,
- all connected via fast networks. And the emergence of many- 1st Edition Morgan Kaufmann, 2011.
core technologies such as GPUs, as well as supercomputers on [9] Ian Foster, Carl Kesselman, and Steven Tuecke, “The Anatomy of the
Grid: Enabling Scalable Virtual Organizations,” The International
chip within these environments has added to the complexity. Journal of High Performance Computing Applications, vol. 15, no. 3,
Thus high performance computing can now use multiple Fall 2001, pp. 200-222.
diverse platforms and systems simultaneously, giving rise to [10] M. Baker, A. Apon, R. Buyya, and H. Jin, “Cluster computing and
applications. In A. Kent & J. Williams (Eds.), Encyclopedia of
the term "computing jungle". Ibis high-performance Computer Science and Technology, pp. 87-125, 2002.
distributed programming system is an example of the jungle [11] Michael A. Rappa, “The Utility Business Model and the Future of
Computing Services,” IBM Systems Journal, vol. 43, no. 1, pp. 32-
computing. 42, 2004.
[12] Microsoft Corporation, “Load-Balanced Cluster”,
http://msdn.microsoft.com/en-us/library/ff648960.aspx [download
Conclusion date: 10 April 2012].
We have discussed the motivation for distributing [13] R. BUYYA, C. S. YEO et al., “Market-oriented Cloud Computing:
computing. It will continue flourishing. There are so many Vision, Hype, and Reality for Delivering IT Services as Computing
Utilities”, in the proceedings of 10th IEEE Conference on High
topics which are going very hot in the research and
Performance Computing and Communications, 2008.
development topics in both the academic and industry for [14] R. BUYYA, C. S. YEO et al., “Cloud computing and emerging IT
many years to come. In above all the cloud computing is the platforms: Vision, Hype, and reality for delivering computing as the
recent topic which is under development by so many 5th utility”, Future Generation Computer Systems, Vol. 25, pp.599-
616, 2009.
industrial giant like Google, EMC, Microsoft, Yahoo,
Amazon, IBM, etc. This review paper enlightens the new
paradigms of distributing computing. It will be beneficial for
the students and the researchers.

ISSN 2278-5299 187

View publication stats