Design and optimization of data center performance

Disk storage technologies have not advanced since the last major introduction of the network storage architecture, mainly in the Storage Area Networks (SAN) and Network Attached Storage (NAS). Hence, it is the main cause of bottlenecks in systems today. Conversely, processing power technologies and...

Full description

Saved in:
Bibliographic Details
Main Author: Tang, Rongjun.
Other Authors: Chng Eng Siong
Format: Final Year Project
Language:English
Published: 2013
Subjects:
Online Access:http://hdl.handle.net/10356/52478
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-52478
record_format dspace
spelling sg-ntu-dr.10356-524782023-03-03T20:37:19Z Design and optimization of data center performance Tang, Rongjun. Chng Eng Siong School of Computer Engineering A*STAR Data Storage Institute (DSI) DRNTU::Engineering::Computer science and engineering Disk storage technologies have not advanced since the last major introduction of the network storage architecture, mainly in the Storage Area Networks (SAN) and Network Attached Storage (NAS). Hence, it is the main cause of bottlenecks in systems today. Conversely, processing power technologies and network technologies have made significant progress to increase their performance. The hybrid storage system is currently widely adopted for data caching, where frequently accessed files will be moved into the SSD to obtain faster I/O, hence contributing to a decrease in overall latency. However, studies have suggested that metadata caching also contribute substantially to a system as it is where all the operations are, thus reducing storage system bottlenecks. In addition, it is more cost-saving. As such, to address whether metadata caching is indeed a beneficial approach in improving storage system bottlenecks, there will be two aims for this study. The first aim is to build on the current literature and determine the contributing factors and relationships pertaining to metadata read and write I/O access during an Online Transaction Process (OLTP). Following that, the second aim of this study is to implement a filesystem with a metadata cache. To achieve both aims, three objectives were identified for each aim. Bachelor of Engineering (Computer Science) 2013-05-09T04:46:20Z 2013-05-09T04:46:20Z 2013 2013 Final Year Project (FYP) http://hdl.handle.net/10356/52478 en Nanyang Technological University 92 p. application/pdf
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic DRNTU::Engineering::Computer science and engineering
spellingShingle DRNTU::Engineering::Computer science and engineering
Tang, Rongjun.
Design and optimization of data center performance
description Disk storage technologies have not advanced since the last major introduction of the network storage architecture, mainly in the Storage Area Networks (SAN) and Network Attached Storage (NAS). Hence, it is the main cause of bottlenecks in systems today. Conversely, processing power technologies and network technologies have made significant progress to increase their performance. The hybrid storage system is currently widely adopted for data caching, where frequently accessed files will be moved into the SSD to obtain faster I/O, hence contributing to a decrease in overall latency. However, studies have suggested that metadata caching also contribute substantially to a system as it is where all the operations are, thus reducing storage system bottlenecks. In addition, it is more cost-saving. As such, to address whether metadata caching is indeed a beneficial approach in improving storage system bottlenecks, there will be two aims for this study. The first aim is to build on the current literature and determine the contributing factors and relationships pertaining to metadata read and write I/O access during an Online Transaction Process (OLTP). Following that, the second aim of this study is to implement a filesystem with a metadata cache. To achieve both aims, three objectives were identified for each aim.
author2 Chng Eng Siong
author_facet Chng Eng Siong
Tang, Rongjun.
format Final Year Project
author Tang, Rongjun.
author_sort Tang, Rongjun.
title Design and optimization of data center performance
title_short Design and optimization of data center performance
title_full Design and optimization of data center performance
title_fullStr Design and optimization of data center performance
title_full_unstemmed Design and optimization of data center performance
title_sort design and optimization of data center performance
publishDate 2013
url http://hdl.handle.net/10356/52478
_version_ 1759857542850150400