Efficient Database Distribution Using Local Search Algorithm

D, Sivakumar and Ch, Sobhan Babu (2011) Efficient Database Distribution Using Local Search Algorithm. Masters thesis, Indian Institute of Technology, Hyderabad.

CS09G004.pdf - Submitted Version

Download (1MB)


A problem in railway database is identied. Focus of the problem is to reduce the average response time for all the read and write queries to the railway database. One way of doing this is by opening more than one database servers and distributing the database across these servers to improve the performance. In this work we are proposing an ecient distribution of the database across these servers considering read and write request frequencies at all locations. The problem of database distribution across dierent locations is mapped to the well studied problem called Uncapacitated Facility Location(UFL) problem. Various techniques such as greedy approach, LP rounding technique, primal-dual technique and local search have been proposed to tackle this problem. Of those, we are using local search technique in this work. In particular, poly- nomial version of the local search approximation algorithm is used to solve the railway database problem. Distributed database is implemented using postgresql database server and jboss appli- cation server is used to manage the global transaction. On this architecture, database is distributed using the local optimal solution obtained by local search algorithm and it is compared with other solutions in terms of the average response time for the read and write requests.

[error in script]
IITH Creators:
IITH CreatorsORCiD
Item Type: Thesis (Masters)
Uncontrolled Keywords: TD10
Subjects: Computer science > Big Data Analytics
Divisions: Department of Computer Science & Engineering
Depositing User: Team Library
Date Deposited: 03 Nov 2014 03:36
Last Modified: 24 May 2019 05:43
URI: http://raiith.iith.ac.in/id/eprint/599
Publisher URL:
Related URLs:

Actions (login required)

View Item View Item
Statistics for RAIITH ePrint 599 Statistics for this ePrint Item