Evaluating paas scalability and improving performance using scalability improvement systems

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 03 Issue: 03 | Mar-2014, Available @ http://www.ijret.org 598
EVALUATING PAAS SCALABILITY AND IMPROVING PERFORMANCE
USING SCALABILITY IMPROVEMENT SYSTEMS
Nishant Agnihotri1
, Aman Kumar Sharma2
1
Assistant Professor, Department of Computer Science & Application, Indus International University, Himachal Pradesh,
India
2
Associate Professor, Department of Computer Science, Himachal Pradesh University, Himachal Pradesh, India
Abstract
Cloud computing has almost changed the way of obtaining resources and managing platform as a service. With these improvements,
challenges like scalability testing, performance testing is emerging very fast. This paper focuses on the evaluation of scalability of
PaaS as well along with that propose a graphical model where SIS (Scalability Improvement System)is proposed in which cache are
implemented at the both interfaces of PaaS to enhance the scalability in terms of providing service to the user.
Keywords: Evaluating PaaS scalability, scalability testing, proposed graphical model of SIS.
--------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODUCTION
Starting with the evolution in the field of technology the
generations comes first in mind. Cloud computing [8, 1, 5] is
the result of these advancements in the technologies or the
generation in the field of technology. In the very first stage
when there were large computers and the costs were very high
the results were not satisfactory but were costly. With the
enhancement in the technology the industry/technology moved
toward the IC’s and the computation time was reduced to a
large extent with the enhancement in the storage capacity.
After these stages the next stage was the “cloud computing”
where the computation cost and storage were given the main
importance. Cloud computing is the result of combination of
various services which are clubbed together to give a massive
change to the new era.
Many technologies clubbed together to give result in form of
cloud computing. These services are grid computing, hybrid
computing, parallel computing, network storage technology
and many other technologies. In short cloud computing gave
birth to the on demand service with the help of some laid
infrastructure. User can demand for any service it can be
software, storage, or any other internet service.
In order to implement cloud computing [8] a basic thing
required is the Infrastructure/Architecture which is
combination of various services.
 IaaS (Infrastructure as a Service): the very first
component/layer of the architecture where the user gets a
platform of a computer system with the help of internet.
 PaaS (Platform as a Service): the second component/layer
of the architecture where a platform for the development of
the software is given to the user though internet.
 SaaS (Software as a Service): the third component/layer of
the architecture where a user can use any software as per
the need through the internet.
Fig1: Cloud computing infrastructure /architecture
Most of the research works in the last few years have focused
toward the performance evaluation and scalability analysis.
Hence, this research paper has a purpose to support evaluation
of performance and scalability analysis for cloud based
applications provides platform for development such as PaaS
(Platform as a Service), SaaS (Software as a Service) and IaaS

__________________________________________________________________________________________
(Infrastructure as a Service) because of some distinguished
features.
 Providing the scalability to the applications based on the
cloud is the major advantage and a distinguished feature.
Assuring this claims the basic and the important issue, and
can be assured by the evaluation of performance
parameters.
 All SaaS (Software as a Service) in cloud are operated with
the physical and virtual computing resources with the
support of scalable infrastructure.
 Using PaaS (Platform as a Service), with the use of
efficient virtualized computing resources that can be scaled
up and down rapidly in a secure way to deliver a high
quality of service new applications can made available [12]
1.1 Research Objective
 The objectives of the research paper are formulated as
under:
 Identifying the scalability factors on PaaS.
 Analyzing the various problems scaling down the
Scalability.
 Improve the efficiency of cloud network with the
implementation of Scalability Improvement System.
1.2 Literature Review
Scalability testing [11, 12, 4] and performance evaluation in
traditional and cloud based applications
There are four primary objectives in evaluating PaaS for the
scalability.
 Check whether the resources are properly allocated to
ensure delivery of service to end user.
 Identify whether the performance parameters are analyzed
properly in order to ensure the connectivity.
 Properly analyze the performance of end system in term of
accessing resources.
 Properly analyze the scalability PaaS at the cloud in order
to assure the quality of service.
There is a big gap and difference in evaluating the scalability
or performance of a traditional and cloud based application
[8]. There are various factors which are helpful in order to
analyze those gaps as the major concern in testing the cloud
based application [7] is fully based on the scalability as well as
the performance in term of delivery of service.
Whereas in traditional software [13] there are some more
factors which are bit different than cloud. Some more factors
are discussed below
 All the applications which works on the cloud environment
provide a contract to the customers known as the SLA [6]
(Service Level Agreement) which focuses on the quality of
service, as the quality is the major part of the cloud
computing in that case this agreement is thoroughly
verified and validated by the Provider of SaaS, PaaS and
IaaS.
 Cost of using data by customer should not more than the
service used.
 Testing in cloud need not to buy any specific software for
the testing whereas in traditional testing this becomes an
unaffordable cost as cloud provides an infrastructure that is
capable of performing testing on various software on single
platform.
 It becomes difficult to create test environments as the
traditional software depends upon various parameters and
configurations as compared to the cloud. These parameters
are, enterprise level firewall configuration, network
configuration and registry settings. Combination of these
parameters grows by the geometry [8], but the resultant
configuration is difficult to manage.
Testing itself playing a common role in both the traditional
testing but still there are few factors which came out when
testing is done in cloud as the variety of testing plans and
Schemes are provided by the provider of the service. Benefits
of cloud testing [8] over traditional testing are as follows
 Cloud incurs financial difficulty: it is an economical aspect
of cloud computing where user pays as per the use i.e.
the amount of data used the proportional will be the charge
[10]. This overcomes the financial difficulty from the
consumer point of view.
 Cloud saves the jitter: it is an effectiveness improving
aspect of cloud computing where the services are assured
to be the fastest one. Delays are minimized due to
implementation of new technologies as it is fully network
driven approach so it improves the packet transfer speed.
 Clouds are easy to expand: it is a fully manageable network
driven approach which ensures scalability. It is easy to
expand by adding new devices to the network resulting
into an easily manageable and expandable network.
 Cloud respond rapidly: In traditional testing where testing
environment is created for different software and different
test cases are designed. Cloud testing saves the time money
and respond time by providing virtualized environment to
the consumer. For the implementation of the testing with
the number of test cases as much work is finished earlier
like implementation of hardware and software installation,
environment maintenance, middleware configuration. And
because of the existing environment respond is given
immediately.
 PaaS scalability: Platform as a Service acts as an
intermediate layer in the cloud computing our idea
approaches toward the various factors for the scaling up as
well as scaling down various components in PaaS. These
factors are mentioned below.
 Factors considered to scale up: evaluating the platform for
various developments, whether it provides the compatible
platform for the rising demand of new technologies.

__________________________________________________________________________________________
 Factors consider to scale down: identifying the various
factors responsible for the scaling down the scalability
even after the load is been removed from the network.
 Cost for the scalability: identifying the various cost factors
for the utilization of resources responsible for the
scalability improvement.
2. PROPOSED MODEL FOR THE SCALABILITY
IMPROVEMENT
This section explains the brief description about the cost factor
in PaaS where economical aspect of cloud computing where
user pays as per the use is considered primarily. A mode is
proposed(Figure 2) based on the old networks where up
gradation with the new devices will be much costly rather than
cutting down the cost of up gradation with in implementation
of a cache scheme on the PaaS layer. As discussed by C.Baker
[2] in his research about the enhancing the scalability by
adding new manageable switches to the network where some
new devices were added to the network. In order to scale down
the cost of the use and improving scalability this step was a
successful attempt. But on the other hand the cost of
replacement was still a major issue. Whereas
T.Banditwattanawong [11] proposed a model where a cache
on the client-side is implemented which was a successful
attempt to improve the scalability and reduction of cost in
terms of SaaS(Software as a Service), but on the cloud side the
scalability in terms of PaaS and IaaS was still a major issue.
SIS (Scalability Improvement System): SIS is a combination
of Load balancer, Load analyzer and caches. Load analyzer;
analyze all the requests which are made by the user to the
system directly or via SaaS. It will keep a track of repeatedly
requested data and saves repeatedly requested elements on the
cache 1 or cache 2 depending upon the interface whether
connected to SaaS or IaaS. When the same item is requested
next time it will be provided effectively by the respective
cache. On the other hand when any of the cache overflows,
Load balancing mode will get activated and transfer the
overloaded data to the other cache resulting in the fast access
to the repeatedly requested elements. Figure (2) shows the
proposed system and its implementation with the architecture
of Cloud computing. Whereas Figure (3) represents the
internal components of the SIS (Scalability Improvement
System) is responsible for the effective functioning of the
system.
Fig 2: Proposed Scalability Improvement system
Fig3: Components of Proposed Scalability Improvement
system
 Cache 1: Cache dedicatedly works for the interface
connected to the IaaS and PaaS. Keep the small objects
with the higher demand from the customer in order to save
the respond time of repeatedly occurring or demanding
item. Cache 1 can work as a load balancing cache to cache
2.
 Cache 2: Cache dedicatedly works for the interface
connected to the SaaS and PaaS. Keep the small objects
with the higher demand from the customer in order to save
the respond time of repeatedly occurring or demanding
item. Cache 2 can work as a load balancing cache to cache
1.
 Load analyzer: The major component of SIS with a control
strategy which will decide the activation of Load
Balancing mode depending upon the size of data on each
cache.
 Load Balancing Mode: Mode of SIS where when any of
cache have load more than other in order to improve the
efficiency this mode will enable and transfer the further
load to the other cache.

__________________________________________________________________________________________
2.1 Functionality & Advantages of Proposed System
SIS works on following steps
1. Assign more contents to the cache which is freer or not in
use: Set a weight on the scale for each cache and the data will
be proportionally distributed on the cache.
2. Route traffic to the dedicated cache till it is available: by
arranging the distribution priority order of data transfer, the
dedicated cache will get the data on first priority. Lower
priority cache will only be used if the current cache gets
overloaded.
3. Give the fastest response time when using applications like
Software Development Kit and more online development
tools.
SIS(C1,C2,LA,REQ)
C1: cache 1
C2: cache 2
LA: Load analyzer (Analyze the load on
cache 1 & cache 2)
REQ: Request (item) (Request for the item by the client)
Step 1: START
Step 2: LA REQ(ITEM)
Step 3: IF REQ(ITEM)=REQ(IaaS)&&ITEM==LA(ITEM)
&& size(C1)> REQ(ITEM)
Step 4 : LA=REQ(ITEM), C1 REQ( ITEM)
Step 5: ELSE IF REQ(ITEM)=REQ(SaaS)&& ITEM==
LA(ITEM)&& size(C2)> REQ(ITEM),
Step 6: C2 REQ(ITEM)
Step 7:ELSE ACTIVE LB.
Step 8: STOP
Steps followed by Load balancing
LB: load balancer (balances the load of the cache when load
balancing mode is active)
LB(C1,C2) (load balancer , gets active when cache 1 or cache
2 overflows)
Step 1: ANALYZE C1&C2
Step 2: IF SIZE(C1)>SIZE(C2)
Step 3: TRANSFER REQ(ITEM) C2
Step 4: ELSE TRANSFER REQ(ITEM) C1
Step 5: STOP
Advantages of Scalability Improvement System:
 Scaling down delays: After implementation of the cache on
the layers responding to the repeatedly demanded data will
be improved. Load balancing scheme provides the backup
to the cache resulting in the improvement of the scalability.
As a result the delays will get scaled down.
 Scaling down the usages cost: As a result of improved
scalability and the fast responding property, used needs not
to wait for the elements for a long time, with this the time
is saved which also improves the cost factor. As a result
the usage cost scaled down up to a good extent.
 Scaling up the connectivity: After implementing SIS
(Scalability Improvement System) connectivity between
the layers become double, requests waiting time will get
scale down up to a good extent, which results in good
connectivity.
 Scaling down the implementation cost: In case of
improving the scalability the preference are given to
change of topology, replacing switches or various servers
which scales up the cost of implementation. Whereas the
proposed SIS can work upon any pre implemented
topology and devices which
3. CONCLUSIONS
With the emergence of new technologies on the cloud testing
and scalability have become a tedious job. There are number
of other methods which work for the testing purpose of
application/web applications/development testing the
scalability and for its improvement. The latest/emerging
techniques offer new setup rather than improvement of the
traditional setups/devices. Although scalability testing and its
improvement is the major issue but replacement of traditional
systems is not the solution at all. SIS is useful for the
traditional/current setups on the cloud. SIS enhances the
scalability along with the reduction in the cost of use of the
network. Future work involves controlling of overflow of the
requests by the customer for multiple items at a single time if
occurs in the Scalability Improvement System and the
implementation of the testing on Scalability Improvement
System.
REFERENCES
[1]. Atif Farid Mohammad and Hamid Mcheick, ”Cloud
Service Testing: An Understanding”, The 2nd International
Conference on Ambient Systems, Network and Technologies,
ELSEVIER, 2012, pp 513-520.
[2]. C.Baker, A.Anjum, R.hill, N.bessis and
S.L.Kiani,”Improving cloud datacenter scalability, Agility and
performance using open flow”, IEEE 4th International
Conferences on Intelligent Networking and Collaborative
Systems, 2012, pp 20-27
[3]. D.Chappell,”A Short introduction to Cloud Platform”
David Chappel and Associates, 2008
[4]. D.Jayasinghe, S.Malkowski, J.li, Q.Wang, Z.Wang, And
Calton Pu, “Variation In Performance and Scalability: An
Experimental Study in IaaS Clouds using Multi-Tier
Workloads”, IEEE TRANSACTIONS ON SERVICE
COMPUTING, IEEE, 2013, pp 1-14.
[5]. G.Gruman” What Cloud Computing really Means”,
InfoWorld, 2009
[6]. J.Gao, P.Pattabhiraman, X. Bai, W.T.Tsai, “SaaS
Performance and Scalability Evaluation in Clouds”, 6th
International Symposium on Service Oriented System
Engineering, IEEE, 2011, pp 61-71.
[7]. M.A.vouk, “Cloud Computing-issues, Research and
Implementation [J]”, Journal of Computing and Information
Technology, CIT 16, 2008 (4), pp 235-246.

__________________________________________________________________________________________
[8]. Peng Zhelonlong ,Ou Yang Zhonghuli and Huang Youlan,
”The Application and Development Of Software Testing in
Cloud Computing Environment”, IEEE International
conference on Computing Service System, 2012, pp 450-454.
[9]. R.Buyya, Y.S.Chee and V. Srikumar,”Market-Oriented
Cloud Computing: Vision, hype and Reality for Delivering IT
Services as Computing Utilities”, Department of Computer
Science and Software Engineering, University of Melbourne
Australia, 2008, pp 9.
[10]. T. Banditwattanawong,”From Web Cache to Cloud
Cache”, 7th International Conf In Grid and Pervasive
Computing, Hong-Kong, 2012, pp 1-15
[11]. Thepparit Banditwattanawong and Putchong Uthayopas,
“Improving cloud scalability, economy and responsiveness
with client-side cloud cache” IEEE, 2013.
[12]. Trieu C. Chieu, Ajay Mohindra and Alexei A. karve,
”Scalability and Performance of Web Application in a
Computing Cloud”, 8th International Conference on e-
business Engineering, IEEE 2011, pp 317-323.
[13]. Zhao Bin, “Use of Senior Software Test Engineer—
classic Tutorial of Software Testing Technology”, [M] Science
Press, 2007, pp- 29-46.

Evaluating paas scalability and improving performance using scalability improvement systems

More Related Content

What's hot (18)

Viewers also liked (17)

Similar to Evaluating paas scalability and improving performance using scalability improvement systems (20)

More from eSAT Journals (20)

Recently uploaded (20)

Evaluating paas scalability and improving performance using scalability improvement systems