Text-only Preview

VAGHELA ANKITA* et al ISSN: 2319 - 1163
Volume: 2 Issue: 5 760 - 763


Vaghela Ankita
PG student, Department of Computer Engineering, Alpha College of Engineering and Technology, Gujarat, India,
[email protected]

Cloud computing is bringing a revolution in computing environment replacing traditional software installations, licensing issues into
complete on-demand services through internet. In Cloud computing multiple cloud users can request number of cloud services
simultaneously. So there must be a provision that all resources are made available to requesting user in efficient manner to satisfy
their need. Resource allocation is based on quality of service and service level agreement. In cloud computing environment, to
allocate resources to the user there are several methods but provider should consider the efficient way to guarantee that the
applications' requirements are attended to correctly and satisfy the user's need This paper survey different resource allocation
policies used in cloud computing environment.

Keywords: Cloud computing, Resource allocation
applications in the cloud. The operating systems and network
access are not managed by the consumer, and there might be
Cloud Computing is a technology that uses the internet and
constraints as to which applications can be deployed. In
central remote servers to maintain data and applications.
Infrastructure as a service Consumers control and manage the
Cloud computing allows consumers and businesses to use
systems in terms of the operating systems, applications,
applications without installation and access their personal files
storage, and network connectivity, but do not themselves
at any computer with internet access. This technology allows
control the cloud infrastructure.
for much more efficient computing by centralizing data

storage, processing and bandwidth.
Resource Allocation

Cloud computing is a model for enabling ubiquitous,
Resource allocation [8] is a subject that has been addressed in
convenient, on-demand network access to a shared pool of
many computing areas, such as operating systems, grid
configurable computing resources e.g., networks, servers,
computing, and datacenter management. A Resource
storage, applications, and services that can be rapidly
Allocation System (RAS) in Cloud Computing can be seen as
provisioned and released with minimal management effort or
any mechanism that aims to guarantee that the applications'
service provider interaction.
requirements are attended to correctly by the provider's

infrastructure. Along with this guarantee to the developer,
Cloud computing customers do not own the physical
resource allocation mechanisms should also consider the
infrastructure; rather they rent the usage from a third party
current status of each resource in the Cloud environment, in
provider. They consume resources as a service and pay only
order to apply algorithms to better allocate physical and/or
for resources that they use.
virtual resources to developers' applications, thus minimizing

the operational cost of the cloud environment.
Cloud computing provide three types of services [7], including

software as a service (SaaS), platform as a service (PaaS) and
Cloud resources can be seen as any resource (physical or
infrastructure as a service (IaaS). In Software as a service
virtual) that developers may request from the Cloud. For
consumers purchase the ability to access and use an
example, developers can have network requirements, such as
application or service that is hosted in the cloud. A benchmark
bandwidth and delay, and computational requirements, such as
example of this is Salesforce.com, as discussed previously,
CPU, memory and storage. Generally, resources are located in
where necessary information for the interaction between the
a datacenter that is shared by multiple clients, and should be
consumer and the service is hosted as part of the service in the
dynamically assigned and adjusted according to demand. It is
cloud. In Platform as a service Consumers purchase access to
important to note that the clients and developers may see those
the platforms, enabling them to deploy their own software and
finite resources as unlimited and the tool that will make this
IJRET | MAY 2013, Available @ http://www.ijret.org/

VAGHELA ANKITA* et al ISSN: 2319 - 1163
Volume: 2 Issue: 5 760 - 763

possible is the RAS. The RAS should deal with these
does the multiplication the total power consumed during the
unpredictable requests in an elastic and transparent way. This
decision making time. It selects only one tier for servers. It
elasticity should allow the dynamic use of physical resources,
finds the ON servers. It calculates the allocated memory for
thus avoiding both the under-provisioning and over-
clients. The profit maximum problem is solved using upper
provisioning of resources.
bound on the total profit. The force between clients and

servers are the major factor in the resource consolidation using
There are different resource allocation policies that are used in
force directed search algorithm. This algorithm takes a client
cloud computing environment. Each of this policy uses certain
based on the highest force towards a new server. It performs
methods and algorithms which are given below:
the load replacement, if the server is available. The algorithm

performs the updating of forces between the clients and
servers. If there is no positive force differential, then the
algorithm stops its working. The algorithm saves the best
solution in each step.
A time-driven adaptive mechanism for cloud resource

Multi-dimensional Resource Allocation Algorithm in
allocation [1]
cloud Computing [3]
Cloud computing service providers deliver their resources
based on virtualization to satisfy the demands of users. In
Cloud computing has emerged as a new technology and it has
cloud computing, the amount of resources required can vary
been increasingly adopted in many areas including science and
per user request. Therefore, the providers have to offer
engineering as well as business. How to arrange large-scale
different amounts of virtualized resources per request. To
jobs submitted to cloud in order to optimize resource
provide worldwide service, a provider may have data centers
allocation and reduce cost is an issue of common concern.
that are geographically distributed throughout the world.
Paper present are two common ways to optimize resource
Likewise, the user locations vary in geographic location. Since
utilization. One is at the application level when applications
cloud computing services are delivered over the internet, there
are arriving, other is in the period of applications running. In
may be undesirable response latency between the users and the
this paper, author makes effort on the former way to address
data centers. Hence, for the best service, the provider needs to
multi-dimensional resource allocation problem by proposing a
find a data center and physical machine that has a light
resource allocation scheme using fewer nodes to process
workload and is geographically close to the user. The
applications. To address multi-dimensional resource
proposed model finds the best match for the user requests
allocation problem, raises several concerns. Firstly, allocate
based on two evaluations: 1) the geographical distances
method should decide which virtual machines should be
between the user and data centers and 2) the workload of data
assigned with a new set of jobs. Secondly, there exists an
centers. Hence, the model allows the users to find a data center
optimal set of nodes which can process new arriving
that is guaranteed to be the closest distance and have the
applications, how to find an efficient way to assign
lightest workload. Also, it finds a light workload physical
applications in nodes is another issue should be solved. In
machine within the data center for a provider.
response to these issues, we use virtual machine as the

minimum resource allocation unit. When a new batch of
Multi-dimensional SLA-based Resource Allocation for
applications arrives, applications is decompose into several
types of jobs, each job with the same type has the same
Multi-tier Cloud Computing Systems [2]
requirement of resource. Aiming at the first issue mentioned
above, author formulate it by adding multi-dimensional
Resource allocation is the most important challenges in cloud
constraints in resource allocation process, which assure jobs
computing. The service provider should work hard for
can be processed in nodes selected by object function. Aiming
allocating resources based on the client's SLA (Service Level
at the second problem, under the purpose of optimize utility of
Agreement).Force directed search algorithm is the solution for
nodes, this nodes select problem as a binary integer
SLA based resource allocation problem for multi-tier
programming problem, and our object function can assure
applications in cloud computing .This algorithm considers the
using working nodes' remain resources to process more jobs.
Gold SLA, and Bronze SLA. The provider gives the guarantee
Unlike the existing resource allocation schemes which allocate
for the response time in Gold SLA. The requests are moved
more physical resources (CPU, memory, etc.) to the exiting
forward and backward in multi-tier service model. The server
virtual machine, this model assign jobs in running nodes to
serves the backward requests. Probability Distribution
make nodes work at a higher utilization.
Function (PDF) is used for finding the arrival rate in the Gold

SLA. The resource management problem's aim is to maximize
the total profit. The profit maximization problem has these
steps. It performs the summation of client's utilities. It
calculates the operation cost of the services. This problem
IJRET | MAY 2013, Available @ http://www.ijret.org/

VAGHELA ANKITA* et al ISSN: 2319 - 1163
Volume: 2 Issue: 5 760 - 763

SLA-based Resource Allocation for Software as a

violations based on the dynamic allocation of resources to
Service Provider (SaaS) in Cloud Computing

Environments [4]
Adaptive Resource Allocation for Pre-empt able Jobs
SaaS is a software delivery method that provides access to
in Cloud Systems [5]
software and its functions remotely as a Web-based service. It
In this paper authors propose an adaptive resource allocation
allows organizations to access business functionality at a cost
algorithm for the cloud system with preempt able tasks in
typically less than paying for licensed applications since SaaS
which algorithms adjust the resource allocation adaptively
pricing is based on a monthly fee. In order to deliver hosted
based on the updated of the actual task executions. Adaptive
services to customers, SaaS companies have to either maintain
list scheduling (ALS) and adaptive min-min scheduling
their own hardware or rent it from infrastructure providers.
(AMMS) algorithms are use for task scheduling which
This requirement means that SaaS providers will incur extra
includes static task scheduling, for static resource allocation, is
costs. Though the cost of the resources has to be minimum, it
generated offline. The online adaptive procedure is use for re-
is also important to satisfy a minimum service level to
evaluating the remaining static resource allocation repeatedly
customers. Saas providers are able to manage the variety of
with predefined frequency. In each reevaluation process, the
customers, mapping customer requests to infrastructure level
schedulers are re-calculating the finish time of their respective
parameters and considering heterogeneity of Virtual
submitted tasks, not the tasks that are assign to that cloud.
Machines. The allocation method uses two different

Policy based resource allocation in IaaS cloud [6] Most of the
ProfminVmMinAvaiSpace. First algorithm is designed to
Infrastructure as a Service (IaaS) clouds use simple resource
minimize the number of VMs by utilizing already initiated
allocation policies like immediate and best effort. Immediate
VMs. The criterion for reusing VM is, it should have
allocation policy allocates the resources if available, otherwise
maximum available space. The algorithm optimizes the profit
the request is rejected. Best-effort policy also allocates the
by minimizing number of initiated VM. Moreover, it
requested resources if available otherwise the request is placed
minimizes number of violations caused by service upgrade
in a FIFO queue. It is not possible for a cloud provider to
because VM has the maximum available space. In such a way,
satisfy all the requests due to finite resources at a time. Haizea
it reduces the penalty caused by upgrading service. The
is a resource lease manager that tries to address these issues by
disadvantage of this algorithm is that it can decrease the profit.
introducing complex resource allocation policies. Haizea uses
The maximum available space is occupied by small number of
resource leases as resource allocation abstraction and
accounts and it leading other requests to be served by a new
implements these leases by allocating Virtual Machines
VM. To overcome the disadvantages of this algorithm,
(VMs). Haizea supports four kinds of resource allocation
reducing the space wastage by using minimum available space
policies: immediate, best effort, advanced reservation and
(MinAvaiSpace) Strategy instead of MaxAvaiSpace Strategy.
deadline sensitive. Proposed dynamic planning based
When there are more than one VM with same type, deployed
scheduling algorithm is implemented in Haizea that can admit
with the same product type as customer request required, the
new leases and prepare the schedule whenever a new lease can
VMs with enough available space to serve are selected. Then
be accommodated. Experiments results show that it maximizes
request is scheduled to the machine with the minimum
resource utilization and acceptance of leases compared to the
available space in a best-fit manner). The proposed algorithms
existing algorithm of Haizea.
minimize the SaaS provider's cost and the number of SLA

adaptive Finding work load of data

cloud center
Response time
Better response time and
resource allocation
between user and data
resource utilization
Multi-Dimensional SLA-

based resource allocation Force
search Service Level Agreement

Cloud algorithm
Maximize total profit
Computing Systems
Multi-Dimensional Multi-Dimensional

allocation Resource
allocation Use
to Utilization and reduce cost
cloud algorithm
allocate resource
of data center
IJRET | MAY 2013, Available @ http://www.ijret.org/

VAGHELA ANKITA* et al ISSN: 2319 - 1163
Volume: 2 Issue: 5 760 - 763


allocation for Software as ProfminVmMaxAvaiSpace Service Level Agreement Minimize
Service Provider and
and Cost
provider's cost and the
Cloud ProfminVmMinAvaiSpace
number of SLA violations
Computing Environments
Resource Adaptive list scheduling

Allocation for Pre-empt (ALS) and adaptive min-
Service Level Agreement
Cloud min scheduling (AMMS)


resource Immediate,
effort, Service Level Agreement
allocation in IaaS cloud
advanced reservation and
deadline sensitive


Jiayin Li, Meikang Qiu, Jian-Wei Niu, Yu Chen,
Zhong Ming, "Adaptive Resource Allocation for Pre-empt
Cloud computing is the delivery of computing as a service
able Jobs in Cloud Systems," in 10th International Conference
rather than a product, whereby shared resources, software and
on Intelligent System Design and Application, Jan. 2011, pp.
information are provided to users over the network. It is an on
demand service because it offers dynamic flexible resource
Amit Nathani, Sanjay Chaudharya, Gaurav Somani,
allocation for reliable and guaranteed services in pay as-you-
"Policy based resource allocation in IaaS cloud", Future
use manner to public. This paper discusses various resource
Generation Computer Systems 28 (2012) 94-103 doi:
allocation policies based on service level agreement and
10.1016/ j.future. / 2011.05.016
quality of service. All above methods used different
Introduction to Cloud Computing, White paper,
techniques for allocating resource to the user. Time driven
Dialogic, Making Innovation Thrive.
gives better response time and increase resource utilization.
Glauco Estacio Goncalves, Patricia Takako Endo,
Multidimensional SLA based algorithm save resources and
Thiago Damasceno, Resource allocation in clouds: Concepts,
increase resource utilization. Multidimensional algorithm
Tools and Research Challenges.
increase resource utilization and reduce cost of data

center.SLA-based policy minimizes the saas provider's cost
and the number of SLA violations. Adaptive resource
allocation policy increase resource utilization. Policy based
Vaghela Ankita (M.E), Alpha College of Engineering and
resource allocation maximize resource utilization.
Technology, Khatraj, Kalol

Kwok; MINJIE Zhang., A TIME-DRIVEN Adaptive
Mechanism For Cloud Resource Allocation. Proceedings of
IC-BNMT2011, 2011, PP441-446
Hadi Goudarzi, Massoud Pedram, Multi-dimensional
SLAbased Resource Allocation for Multi-tier Cloud
Computing Systems, in Proceedings of IEEE International
Conference on Cloud Computing (CLOUD),Washington DC
USA, 2011.
Bo Yin, Ying Wang, Luoming Meng, Xuesong Qiu,
A Multi-dimensional Resource Allocation Algorithm in Cloud
Telecommunications, Beijing 100876, China, 2012
Linlin Wu, Saurabh Kumar Garg and Rajkumar
Buyya, 2011. "SLA-based Resource Allocation for Software
as a Service Provider (SaaS) in Cloud Computing
Environments". 11th IEEE/ACM International Symposium on
Cluster, Cloud and Grid Computing, Pages 195-204.
IJRET | MAY 2013, Available @ http://www.ijret.org/