ISSN (Online) 2278-1021

ISSN (Print) 2319-5940

International Journal of Advanced Research in Computer and Communication

Engineering
Vol. 4, Issue 3, March 2015

Comparative Performance Analysis of MySQL

and SQL Server Relational Database
Management Systems in Windows
Environment
Amlanjyoti Saikia1, Sherin Joy2, Dhondup Dolma3, Roseline Mary. R4
PG Scholar, Department of Computer Science, Christ University, Bengaluru, India 1,2,3
Assistant Professor, Department of Computer Science, Christ University, Bengaluru, India 4
Abstract: The enormous amount of data flow has made Relation Database Management System the most important
and popular tools for persistence of data. While open-source RDBMS systems are not as widely used as proprietary
systems like Oracle DB or SQL Server, but over the years, systems like MySQL have gained massive popularity. In a
stereotypical view, SQL Server is considered to be an enterprise-level tool, MySQL has carved a niche as a backend for
website development. This paper is an attempt to set a benchmark in comparing the performance of MySQL against
SQL Server in Windows Environment. To test and evaluate the performance the Resort Management System named
Repose is considered. The result shows that SQL Server is still a significantly better performer when compared to
MySQL.

Keywords: MySQL, SQL Server, Performance Analysis,Repose.

I. INTRODUCTION resources and the fact that the project happens to use the
Repose is a Resort Management System designed for relational model in the database.
managing the records of a resort. It stores information of
the users and the resort, which includes services available
in the resort, room details, reservation details, et al.
Repose RMS is designed in such a way that the staffs as
well as the guests feel comfortable in using the software.
Repose RMS enables a guest to register themselves online,
as well as register through staff manually. The intention of
this project is to develop a software which is more user
friendly, and can be efficiently used by people in different
roles.

We saw the development of this project as an opportunity

for analysing the comparative performance of MySQL and
SQL Server. The main focus of this paper is to analyse the
performance of the system in two databases namely - SQL
Server and MySQL and to discover which database is well
suited to work with this system.

II. WHY MYSQL AND SQL SERVER?

MySQL and SQL Server are two of the most popular
RDBMS systems. SQL Server is the most used database
system in organizations, while MySQL is the third most
popular [4]. In overall-use rankings however, MySQL is
the second most popular database system after Oracle DB,
while SQL Server is the third most popular [5]. The
mismatch in these rankings is considered because SQL
Server is seen more of an enterprise tool while MySQL is
considered a tool that appeals most often to individuals
interested in managing databases associated with their
websites [6].
MySQL and SQL Server as the databases were selected
based on the convenience of the developers, available
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.4339 1
 ISSN (Online) 2278-1021
ISSN (Print) 2319-5940

International Journal of Advanced Research in Computer and Communication

Engineering
Vol. 4, Issue 3, March 2015
SQL Server is Microsoft's relational database management
A. MySQL system (RDBMS). It is a fully-featured database which is
MySQL is the world’s second most used Database primarily designed to compete against the likes of Oracle
management system [5], and the most popular of all Database (Oracle RDBMS) and MySQL.
open- source RDBMS systems. It provides many Features of SQLServer:
features, the most valuable of which is its platform  Enables memory optimization of selected tables and
independence. The various features of MySQL and SQL stored procedures.
server are as follows- Features of MySQL:  Provides Migration Assistant programs to migrate
 It can work on multiple platforms. data from the most widely used DB systems.
 Uses Multi-layered server design with independent  Clustering Services that allow to recover instantly
modules. from one system to another.
 Executes very fast.  Replication Services that keep data synchronized in
 Supports many data types. between SQL Server and other DBMS systems.
 Uses a very fast thread-based memory allocation
III. RELATED WORKS
system.
 Supports fixed-length and variable-length records. In [1], the performance comparison for data storage of
health care with two databases namely Dbo4 and MySQL
B. SQL Server was done. The authors have made a detailed study about
data storage in health care and have made an analysis of out the database system that works best for the Repose RMS
data storage in both the databases. The authors have system. As of now the user base of the system is very small
considered both execution time and memory consumption and hence the impact on the back-end is not very huge.
as a parameter for performance analysis. However, knowing how the different database systems are
going to perform beforehand is beneficial in choosing in
In [2], performance comparison between MySQL and between them.
Virtuoso Universal Server 6 triplestores was done. They
have done comparison of these two system on the basis of B. Data Collection
average read and write times. Repose RMS acts as a tool for data collection in this
performance study. The daily interaction of the users with
In [3], performance comparison for row-storage vs the system fills up the database gradually with clean,
column-storage in databases, using SQL Server and Oracle formatted and actual data.
DB as test database system was done. This paper had
analysed the execution time where a sequence of five C. Data Preparation
SELECT queries against database systems were executed, Data cleaning is a process of detecting corrupted and
using MATLAB as a front-end. inappropriate data from a dataset and correcting it. Data
which are inconsistent are removed from the table and are
IV. PROPOSED METHODOLOGY not considered for the performance analysis. Data cleaning

Business
Understanding

Performance Data
Analysis Collection

Deployment
Data Preparation

Fig 1: The test method

A. Business Understanding
As mentioned before, the purpose of this paper is to find
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.4339 2
 ISSN (Online) 2278-1021
ISSN (Print) 2319-5940

International Journal of Advanced Research in Computer and Communication

Engineering
Vol. 4, Issue 3, March 2015
is minimal in this case since the forms take formatted and the database well suited for Repose RMS is found out.
inputs. Even then, some data in the tables are manually
cleaned Fig 2: Design of the test system
D. Deployment
The data access layer in the architecture of the RMS
This system concentrates on the performance analysis of system is responsible for CRUD (Create, Read, Update,
the backend with respect to the databases MySQL and and Delete) operations as well as logging the time taken to
SQL server. The different modules of the system are execute the operations. The Data Access Layer uses the
executed in both the databases and the execution time PerformanceLog class to log entries in an XML file. This
(time taken by the database to return the result of a query) logging operation is done only if Performance Analysis
is recorded in each case. This record is later utilized to mode is set to on in the configuration file.
find out which database is best suitable for this Sample log
application with respect to the time it takes to execute the <Entry TimeStamp="26-Feb-2015 Thursday, 03:51:56 PM
query. IST">
<ConnectionString>odbc:Driver={MySQL ODBC 5.3 UNICODE
E. Performance Analysis Driver};Server=localhost;Database=resortdb;
Uid=user;Pwd=password;</ConnectionString>
The software is executed, using both the databases and <Query Category="SELECT" RowCount="1500"
performance is analysed based on the time it takes to HasJoin="FALSE" HasConditions="FALSE">SELECT * FROM
complete its execution. The execution times are obtained users</Query>
<TimeTaken>0.0053150653839111</TimeTaken>
and then tabulated Microsoft Excel for further processing.
</Entry>
The performance analysis is done on both the databases
These entries provide us with data for the performance INSERT 100 rows was calculated.
comparison between MySQL and SQL Server.
C. UPDATE queries
V. COMPARISON
The comparison test was performed on a Windows 8 64-
bit machine running on an Intel® Core™ i5-2430M CPU
with a clock frequency of 2.4 GHz. and 4GB of RAM.

The SQL Server 2008 and MySQL 5.6.17 were the

respective versions of SQL Server and MySQL to test the
data. The data was first collected in MySQL, and then
migrated to SQL Server using the Microsoft SQL Server
Migration Assistant for MySQL program.

Different DML queries, namely SELECT, INSERT,

UPDATE, and DELETE were executed on both and the
execution time was recorded via the RMS system.

The test cases were as follows –

A. SELECT queries
Four SELECT queries were executed on both SQL Server
and MySQL server – Two UPDATE queries, the first one to update certain rows
that fulfil certain conditions, and the second one to update
 A non-conditional SELECT query all the rows, were executed.
 A SELECT query with an ORDER clause on a non-
The queries were executed five times each, first with 1500
indexed column rows and then with 2000 rows and their average was found
out.
 A SELECT query with a JOIN
D. DELETE queries
 A SELECT query with a JOIN and an ORDER clause
In case of DELETE queries as well, two queries were
on a non-indexed column executed.

All four queries were executed five times with 3000 rows The first one was to remove the last 500 of 1500 rows and
and five times with 5000 rows. the next one to remove all remaining 1000 rows.

B. INSERT queries Both queries were executed five times and the average
On both MySQL and SQL Server, the average time to found out.
Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.4339 3
 ISSN (Online) 2278-1021
ISSN (Print) 2319-5940

International Journal of Advanced Research in Computer and Communication

Engineering
Vol. 4, Issue 3, March 2015
VI. RESULTS

Time (in seconds)

0.1
0.0
9
0.0
8
0.04
0.0
87
7
0.0 0.02
6 61
0.0
5 [VALUE]
0.0 0.0012
4 MySQLMySQLSQLSQL
0.0 (3000(5000ServerServer
3 rows)rows)(3000(5000
0.0 rows)rows)
2
Fig. 3: Averages for a non-conditional SELECT query
0.0
1
Time (in seconds) 0.1
0 0.08
0.0 89
9
0.0
8 0.05 0.05
0.0 69 53
7
0.03
0.0 25
6
0.0
5
0.0
4 MySQLMySQLSQLSQL
0.0 (3000(5000ServerServer
3 rows)rows)(3000(5000
0.0 rows)rows)
2
Fig. 4: Averages for SELECT query with an 0.0
ORDER clause on a non-indexed column 1
Time (in seconds)

0.1 0
0.09 0.0848
0.08
0.07
0.06
0.05 0.0445
0.04
0.03
0.02
0.01 0.0011 0.0011
0
MySQLMySQLSQLSQL
(3000(5000ServerServer
rows)rows)(3000(5000
rows)rows)

Fig. 5: Averages for SELECT query with a JOIN

Copyright to IJARCCE DOI 10.17148/IJARCCE.2015.4339 4

 Time (in seconds) 0.2 0.1

Time (in seconds)

0.18 0.09 0.0857
0.16
0.1221 0.08
0.14
0.12 0.07
0.1 0.0671 0.06
0.08 0.0584 0.0611 0.05
0.06
0.04 0.0311
0.04
0.02 0.03
0 MySQL MySQL SQL SQL 0.02
(3000 (5000 Server Server 0.01
rows) rows) (3000 (5000 0
rows) rows)
MySQL SQL Server

Fig. 6: Averages for SELECT query with a JOIN and an

Fig. 9: Averages for non-conditional DELETE query
ORDER clause on a non-indexed column

0.001

Time (in seconds) 0.0008

Time (in seconds)

0.002
0.0017
0.0006

0.0015
0.0004
0.0011 0.0004 0.0003
0.001 0.0002
0.0002 0.0001

0.0005 0
MySQL MySQLSQLSQL
(1500(2000ServerServer
0 rows)rows)(1500(2000
MySQL SQL Server rows)rows)

Fig. 7: Averages for 100 INSERT queries Fig. 10: Averages for conditional UPDATE query

0.1 0.01
Time (in seconds)

0.016
8 1
0.09
Time (in seconds)

0.01
0.08 6
0.07 0.0655 0.01
4 0.008
0.06 8
0.01
0.05
2
0.04 0.01 0.002
0.002
0.03 0.0219 0.00 4 7

0.02 8
0.00 MySQL MySQLSQLSQL
0.01
6 (1500(2000ServerServer
0 0.00 rows)rows)(1500(2000
MySQLSQL Server 4 rows)rows)
0.00
Fig. 8: Averages for conditional DELETE query Fig. 11: Averages for
2 non-conditional UPDATE query
VII. CONCLUSION 0
The purpose of this paper is to analyse the performance of
two popular relational database management systems, BIOGRAPHIES
MySQL and SQL Server, in terms of time taken to
respond to requests. Sherin Joy is a PG scholar in the
Department of Computer Science, in Christ
The results show that SQL Server offers more University. She secured the first rank in her
performance than MySQL in terms of response time. In BSc degree. She was also awarded first
all the test cases, except INSERT queries, SQL Server class in International Chinthana
consistently took lesser time when compared to MySQL. Mathematics 2004-05. She lists Embedded
Systems and Database Systems as her academic interests.
MySQL also performed poorly in terms of scaling up.
MySQL shows a two-fold increase in time taken when the
number of rows go up. SQL Server also showed similar Dhondup Dolma was born in Tibet. She
results, but the increase in time taken wasn’t as great as earned her Bachelor's degree in Science
MySQL. (Physics, Mathematics and Computer
Science), and is currently pursuing her
The most striking difference in performance was in Master's degree in Computer Science from
SELECT statements. In Fig 3, we can see that the time Christ University. She lists Java and
taken by MySQL is two orders of magnitude more than Network Security among her academic interests as well as
SQL Server when dealing with 3000 rows. travelling and exploration as her passion.

While it may seem that SQL Server is the obvious choice

as a backend, its cost is an obstacle for implementation, Amlanjyoti Saikia is from Assam, where
for a small business. On the other hand, the open-source he did his Bachelor’s degree in Information
nature of MySQL means that implementation costs will through distance education. He is currently
be minimal. MySQL is capable enough to be used as a pursuing his Master’s degree in Computer
backend for a website, and other small-scale applications. Science from Christ University. He lists AI
and robotics among his academic and
In our future work, the increase in the scope of the research interests as well as photography, riding and
analysis in terms of parameters, database systems, DBMS travelling as his passion.
models, and execution environment will be considered.
Also, the comparison of the execution time and memory Prof. Roseline Mary. R currently serves
consumption of RDBMS and NoSQL systems on both as an Assistant Professor in the Department
Windows platform and UNIX platform can be considered of Computer Science, Christ University.
for further enhancement. She obtained her BSc (Mathematics),MCA
from Bharathidasan University, MPhil
REFERENCES from Alagappa University and MTech
[1] S. Kulshrestha and S. Sachdeva, “Performance Comparison for from Karnataka State Open University. Her research
Data Storage - Db4o and MySQL Databases”, Department of interests include Database Systems, Big Data, Data
Computer Science and Engineering, Jaypee Institute of Information Mining and Cloud Computing.
Technology, Noida, India, 2014.
[2] V. Kilintzis, N. Beredimas, and I. Chouvarda, “Evaluation of the
performance of open-source RDBMS and triplestores for storing
medical data over a web service”
[3] A.K. Dwivedi, C.S. Lamba, S. Shukla, “Performance Analysis of
Column Oriented Database versus Row Oriented Database”,
International Journal of Computer Applications (0975 – 8887), Vol.
50- No.14, July 2012
[4] J M. Emison, “2014 State of Database Technology”,
InformationWeek, San Francisco, CA, Rep. R7770314, 2014
[5] DB-Engines.com (2015, March 06). DB-Engines Ranking [Online].
Available: http://db-engines.com/en/ranking/relational+dbms
[6] Lee J. (2013, November 30). Oracle vs. MySQL vs. SQL Server: A
Comparison of Popular RDBMS [Online].
Available: https://blog.udemy.com/oracle-vs-mysql-vs-sql-server/
[7] Y. Bassil“A Comparative Study on the Performanceof the Top
DBMS Systems”, Journal of Computer Science & Research, Vol. 1
- No. 1, February 2012