Comparative Study of REST and SOAP: Case of

Registrar of Political Parties Kenya

Christopher Kiama1, Lawrence Muchemi2
School of Computing and Informatics
University of Nairobi, Kenya
Email: cwkiama@gmail.com1, lmuchemi@uonbi.ac.ke2

Abstract
This research is a comparative study of Representational State Transfers (REST) and Simple Object Access
protocol (SOAP). The research is an objective experimental study that investigates how REST compares to
SOAP web services in terms of performance, scalability, maintainability and extensibility. Two prototypes
one for each web service were developed that consolidate data from different sources into the same
database tables. Data from registrar of political parties of Kenya was presented from MYSQL database via
the web services in the forms of HTML, XML and JavaScript Object Notation (JSON) using the common
browsers.
On average from the experiments performed in this research, Performance results indicated by the amount
of time taken to pull data from the server and the memory required to finish the process, shows that SOAP
web services takes approximately over one and half times memory required by REST and about 3 times the
amount of time needed to fetch the same data from the same database table. However scalability measured
via throughput values (Kbps) against number of users accessing the same data at the same time, indicates
that both REST and SOAP scale almost evenly but with REST having a slightly higher throughput with the
increase in number of users.

Keywords (TXT) and XLS. The two web services will

SOAP, REST, Scalability, web service, help in determining the how well web
performance, extensibility. applications will improve in performance,

1.0 Introduction scalability and extensibility.

In recent decades, systems developed using Representational State Transfer (REST)

web for concurrent information access has marks software architecture pattern (in

been a major issue as far as performance, contrast to SOAP which is a protocol). It

scalability and extensibility are concerned. relies on a stateless, client-server, cacheable

This research which is a case of Registrar of communications protocol and in virtually all

political parties, the body charged with the cases; the HyperText Transfer Protocol

responsibility of registering political parties (HTTP) protocol is used.

before going for elections in Kenya has used REST and SOAP uses a simple HTTP-based

Web service technologies (REST and protocol that enables users to contact the

SOAP) to collect data / information from message broker through a Web browser by

political parties data stores; in the forms of navigating to appropriately formatted

comma separated values (CSV), Text files Uniform Resource Locator (URL) or by

Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 105

posting HyperText MarkUp Language
(HTML) forms based on a subset of the
HTTP protocol [2].
SOAP stands for Simple Object Access
Protocol and brings its own protocol and
focuses on exposing pieces of application
logic (not data) as services. It exposes
operations and is focused on accessing
named operations through different
interfaces.
Though SOAP is commonly referred to as
web services it has very little if anything
to do with the Web but REST provides true
Web services based on URIs and HTTP.
SOAP is not made for resource constrained
mobile devices but it is for fixed network.
The SOAP messages has heavy payload. In
contrast, the REST messaging framework
has lightweight payload which is suitable for
mobile as well as cellular network. REST
identifies the service resources by single
URL only [5].
SOAP request uses POST and require a
complex XML request to be created which
makes response-caching difficult While
RESTful APIs can be consumed using
simple GET requests, intermediate proxy
servers / reverse-proxies can cache their
response very easily.
SOAP consumes more bandwidth because
its response could require more than 10
times as many bytes as compared to REST
Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 106
while REST consumes less bandwidth
because its response is lightweight.
In the development of prototypes using
REST and SOAP, a research is done to
determine the best web service in terms of
scalability, performance, extensibility and
efficiency in solving the main problem of
the registrar of political parties.
Extensibility is the ability of the prototype
system to allow and accept significant
extension of its capabilities, functionality,
enhancement for the purpose of meeting
future needs and significantly changing
requirements without major rewriting of
code or changes in its basic architecture [8].
Scalability is the ability of the prototype
architecture to grow to accommodate
increasing number of users, applications and
system [6].
The objectives of this research are:
1. Design frameworks that Combine
Registrar of political parties data
from different data sources and use
REST and SOAP frameworks to
present it through a browser.
2. Develop prototypes that use
simulated political parties data
based on HTTP web service on
REST and SAOP principles that
demonstrates the framework in
objective one.
 3. Evaluate and compare the in CSV, TXT and XLS formats; the
prototypes based on REST and simulated data is presented in parties and
SOAP frameworks. members tables.
4. Recommend the most suitable
framework for political parties 2.2 Designs frameworks
problem in terms of performance, The research is a comparative study of
scalability extensibility and REST and SOAP; two prototype designs are
maintenance. designed, implemented and comparatively
comparing the results of the experiments
1.1 Scope of the research done.
This research involves designing of frame
works and come up with prototypes that
combine and works with data using REST
and SOAP web services. This will enable
and guide the registrar of political parties in
analyzing and determine presentation of
each political party in every county. The
consolidation of data will be limited to SQL
databases using TXT, CSV files and
spreadsheet formats.
SOAP framework
2.0 Approach
2.1 Data sources and preparation FIGURE 1: MODEL FRAMEWORK OF SOAP
The prototypes are based on political
parties simulated data, which is presented

REST framework

HTTP
Client Service

Option 2: A RESTful service provides a client library Clients see methods with parameters

Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 107

 A
HTTP
Client P Service
I used is configured, server-side code needs to
FIGURE 2: MODEL FRAMEWORK OF REST
2.3 SOAP Implementation
SOAP uses WSDL file that defines the port
which is later mapped to the overarching
Portal SOAP service set. The port itself is
composed of multiple operations, each one
representing a single service to
implemented in the set. In turn, each
operation maps previously-defined SOAP
messages as its input and output types [1].
All the methods and functions in SOAP
server must be registered with WSDL and a
SOAP client created to access the defined
methods. The content of this WSDL file
represents a language- and platform-neutral
method of remotely communicating the
SOAP service interface. Once the WSDL
file for the SOAP implementation of the
service-oriented architecture is written, it
needs to be made publicly available to all
portal clients capable of reaching
particular Portal server. This is done through
the use of a SOAP-enabled HTTP web
server; and may be enabled in a variety of
different ways.
Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 108
Regardless of how the HTTP server to be
be written that is responsible for handling
incoming service requests and formulating
appropriate responses.
2.4 REST Implementation
While SOAP adheres very closely to the
RPC model, REST revolves around the
concept of resources and focuses on using
be the inherent power of HTTP to retrieve
representations of these resources in varying
states. In REST style, every resource is
signified by a unique URL which may be
operated on by a subset of the core set of
HTTP commands: Get, Post, Put, and Delete
[10].
Compared to SOAP, instead of a specialized
SOAP client, any HTTP client library may
be used to interact with the Portal REST
server. Every contemporary language comes
equipped with a built-in HTTP library as it
is a universal protocol for communicating
over the Internet [4].
a
3. Evaluation
Comparatively, from the experiments that
were don in this research; REST has proved
to be better and easy to maintain as it is an
architecture that needs very few lines of
code to do any urgent changes required
during maintenance or modifications to
solve a specific problem compared to SOAP
which requires modification of the WSDL,
client and server files. In terms of
performance, REST takes shorter time to
load similar amount of data as compared to
SOAP and also it uses less resources e.g.
Memory and processing power.
In terms of complexity REST API accesses
data directly from the database tables while
SOAP requires a soap server, WSDL and a
soap client to access the same data from the
database table.
Use of REST frame work will help the
registrar of political parties manage partys
membership with computing resources that
have better performance, maintainability and
extensibility incase of future needs.
a) Experimental procedures
As previously mentioned, the main objective
of this project is to investigate how REST
and SOAP compares and to examine the
performance, scalability, maintainability and
extensibility of the Web Services in
application development.
The experimental procedure followed
evaluates the best web service to be used.
Following is description of the experiments
taken:-
i) Performance measure
Two prototypes are used to carry out the
experiments needed to check on the
performance of the Web Services during
data access from the database server. The
Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 109
two prototypes are used to access the same
data for political and members data from the
registrar of political parties database and the
results are accessed and viewed from a web
application.
The evaluation for both services is carried
out using two different scenarios. In the first
set of experiments the level of internal
resource consumption is examined including
memory resources. In the second set of
experiments the overall performance is
evaluated by measuring total elapsed
response time for execution of each request.
After that, evaluation for the two web
service is done to determine the best
performer [9].
ii). Scalability measure
The two prototypes REST and SOAP are
experimented on how well they grow or
expand with the increased users; this was
done by using apache-jmeter which tests
throughput with the increase in the number
of users accessing the same data.
For the two prototypes you create a test plan
for threads (users), the set the thread
properties, finally you add and set the
requests by specifying the server name (IP
of the server), the path of the HTTP request,
where the listeners will get their results
from.
The process / experiments are repeated as
many times as possible depending on the
number of users (samples) you intend to run.
The data used in the experiment is simulated data from the server and the memory
data that is got by consolidating it from required in finishing the process.
different sources i.e. CSV, TXT and XLS Scalability of the prototype architecture is
into the MYSQL database. The web service determined by how it grows to
technologies will now fetch the data and accommodate increasing number of users,
present it through a common browser. applications and system.
Maintainability and Extensibility of the
b) Parameters to be measured prototype is measured by how it can accept
The main parameters that are measured in significant extension of its capabilities,
this research are performance, scalability, functionality, enhancement for the purpose
maintainability and extensibility. of meeting future needs and significantly
changing requirements without major
Performance of the prototypes is determined rewriting of code or changes in its basic
by the amount of the time they take to pull architecture.
Experimental Results and SOAP web services which are depicted
a. Performance in Figures 1 and 2.
Table 1: below shows the performance
results for time and memory used by REST
Transaction Data REST REST SOAP SOAP
ID Records Memory Time Memory Time
1 10 2.4 0.06 3.504 0.1842
2
40 3.0 0.8 4.38 2.456
3 70 4.0 0.1 5.84 0.307
4 110 4.4 0.26 6.424 0.7982
5 140 5.2 0.3 7.592 0.921
TABLE 1: REST AND SOAP DATA ON PERFORMANCE

Memory Usage Graph

Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 110

FIGURE 3: MEMORY REQUIRED FOR DATA ACCESS IN REST AND SOAP

Figure 3: show RESTful web services need table compared SAOP which needs over one
by far less amount of memory to fetch the and half times of memory to fetch the same
same amount of data from the same database amount of data in bytes.

Time Usage Graph

FIGURE 4: TIME REQUIRED FOR DATA ACCESS IN REST AND SOAP

Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 111

Figure 4: shows the response time in
seconds for accessing the same amount of
data for the two web services. The response
time of RESTful web service is smaller than
for SOAP web service which is three times
more.
Scalability
Table 2: below shows the scalability results
measuring throughput based on the number
of users accessing the same data using
REST and SOAP web services.

USER SAMPLES

Sample (Users) 10 20 30 40 50 60 70 80 90 100

REST Throughput 5.1 36.7 29 27 27 24 24 23 23.2 46

SOAP throughput 10 32.5 14 17 19 21 23 24.9 26.2 27

TABLE 2: REST AND SOAP DATA THROUGHPUT BASED ON USERS

Scalability of SOAP and REST

FIGURE 5: SCALABILITY OF REST AND SOAP MEASURED USING THROUGHPUT

Figure 5: shows the throughput in bytes per both REST and SOAP scale almost evenly
second as the number of users connecting to but with RESTful web service having a
the server increase. The graph indicates that

Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 112

slightly higher throughput with the increase
in number of users compared to SOAP.
4.0 Results Discussion
These results support that RESTful web
service is recommended for web application
development. Therefore, REST offers a
perfectly good solution for the majority of
implementations with greater advantages
and lower overheads as has the best
performance and scalability.
4.1 Implication
Use of REST frame work will help the
registrar of political parties manage partys
membership with computing resources that
have better performance, scalability and
maintainability.
4.2 Impact to the potential users
The usage of web service technologies will
be of great assistance in providing sanity
and a solution to the registrar of political
parties problems of registering genuine and
legitimate members to various political
parties. REST has proved to be very
efficient in data consolidation from different
sources into a common database and data
access with improved performance and
scalability.
4.3 Conclusion
REST and SOAP differ in context and
usage; SOAP is a protocol web service
Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 113
while REST is architecture web service
though they all use common standards.
From the experiments done in this research,
REST has proved to be better in terms of
performance and scalability as well as being
easy to maintain as it is an architecture that
needs very few lines of code to do any
urgent changes required during maintenance
or modifications to solve a specific problem.
In terms of performance, the response time
of RESTful web service is smaller than for
SOAP web service which is three times
more. In addition RESTful web services
need by far less amount of memory to fetch
the same amount of data from the same
database table compared SAOP which needs
over one and half times of memory to fetch
the same amount of data in bytes.
REST has also demonstrated high degree of
scalability with increasing number of users
accessing the system at the same time;
REST has a slightly higher throughput with
the increase in number of users compared to
SOAP web services.
In terms of complexity REST API accesses
data directly from the database tables while
SOAP requires a soap server, WSDL and a
soap client to access the same data from the
database table.
4.4 Future Research
Future research may address security and
the analytical bit of the captured data,
comparing REST and SOAP; including Web Services grows every day, but REST
automatic alerts in case the system and SOAP are probably the most popular.
encounters are member who is registered SOAP being the first technology it was
more than once across different political rapidly becoming the standard protocol for
parties. accessing Web Services using XML
The research should also measure the same messages to exchange information across
benchmarks on machines of different specs endpoints, and provides several advantages
and model comparing REST and SOAP. over other binary protocols.
With advancement of the technology, Web
5.0 Related Work Services are the key point of Integration for

Recent researches have comparatively used different applications belonging to different

SOAP and REST web service technology Platforms, Languages, and systems. To

measuring different aspects on quality of achieve this, Representational State Transfer

service. Measuring quality attributes of web (REST) which basically means that each

services is inherently difficult due to the unique URL is a representation of some

very virtues of service-oriented object was brought into action so as to outdo

architectures. Different aspects of SOAP.

performance measurement and REST APIs havent been around for long

benchmarking of web services have been and their APIs are definitely modern for

analyzed. However, most approaches do not creating most of the web services [6].

provide concrete ways of measuring

performance of services in a specific 5.1 Yahoo Web Services based on
architecture. Detailed performance REST approach
measurement of web services is particularly Yahoo provides a variety of Web services at
important for obtaining quality attributes http://developer.yahoo.net/, from searching
that can be used for service selection and the web to interfacing with Flickr (Yahoos
composition, and for discovering photo-sharing community). Using REST,
bottlenecks to enable optimization of you can easily add integration for them
composite service processes [8]. within a Web page or larger application.
Although several services are available, this
Web Services is a modern and popular case study demonstrates how to perform a
technology in managing institutional data Web search
and systems. To manage the Web service a (http://developer.yahoo.net/search/web/) and
list of protocols and technologies related to how to perform a product search

Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 114

(http://developer.yahoo.net/shopping/V1/pro
ductSearch.html). Using the ideas and
techniques presented in the examples, it is
quite easy to apply them to access other
offered services.
The Flickr service requires its own API key.
Although a lot of functionality does not
require authentication, to upload photos you
must also register for an API secret key (at
http://www.flickr.com/services/api/registere
d_keys.gne) once you have your API key.
5.2 Use of SOAP in the Mobile
Technology
SOAP has been used in Unstructured
Supplementary Services Data (USSD)
service, which allows high speed interactive
communication between mobile subscribers
and applications across Interfaces to SOAP,
XML, LDAP services
Basically the way USSD works is that you
send some cryptic number like *333# to
your mobile servers who in turn forward the
request via various protocols mostly HTTP
or SOAP but not limited to these.
5.3 Acknowledgment
My sincere gratitude goes to Mr. Lawrence
Muchemi, Dr. E. Opiyo, Mr. A. Mwaura,
Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 115
Mr. C. Ronge for their time, input and also
for their contribution in critiquing the paper.
References
[1] Chappell, D. (2009). SOAP vs.
REST:Complements or Competitors?
San Francisco, California: Chappell &
Associates.
[2] Elkstein, M. (n.d.). Learn REST: A
Tutorial. Retrieved February 1, 2013,
from Learn REST: A Tutorial:
http://rest.elkstein.org/
[3] Gavin Mulligan, D. G. (2009). A
Comparison of SOAP and REST
implementations of a service based
interaction independence middleware
framework. Proceedings of the 2009
Winter Simulation Conference , 3-6.
[4] Jon, F. (2007, July). Which is better,
REST or SOAP? Retrieved July 11,
2013, from SOAP, REST, and More:
http://msdn.microsoft.com/en-
us/magazine/dd942839.aspx
[5] Kishor Wagh, R. T. (2012). A
Comparative Study of SOAP Vs REST
Web Services Provisioning.
Journal of Information Engineering
and Applications , 2-4.
[6] Singh, T. (2009, August 24). REST vs.
SOAP The Right WebService.
Retrieved January 28, 2013, from
REST vs. SOAP The Right
WebService: http://geeknizer.com/rest-
vs-soap-using-http-choosing-the-right-
webservice-protocol/#ixzz2IcJgJ2yO
[7] wuher. (2010, November 12). Rest vs.
Soap. Has REST a better performance?
Retrieved July 11, 2013, from web
services - Rest vs. Soap. Has REST a
better performance? - Stack Overflow:
http://stackoverflow.com/questions/416
3066/rest-vs-soap-has-rest-a-better-
performance
[8] Christoph Becker, H. K. (2009). An
Extensible Monitoring Framework
forMeasuring and Evaluating Tool
Performance in a Service-oriented
Architecture. 1-3.
[9] Feda AlShahwan, K. M. (2010).
Evaluation of Distributed SOAP and
RESTful Mobile Web Services.
International Journal on Advances in
Networks and Services, vol 3 no 3 & 4,
year 2010, , 8-10.
[10] Maven. (2011). Lightweight REST
framework. Retrieved June 4, 2013,
from Lightweight REST framework:
http://essentialsource.sourceforge.net/d
ocumentation/rest.html

Book Chapter: Trends in Distributed Computing Applications-978-9966-074-13-3Page 116