Optimization of Large Scale of Files Transfer in Meteorological Grid Optimization of Large Scale of Files Transfer in Meteorological Grid

被引：0

作者：

Ma, Tinghuai ^{[1
]}

Cao, Hao ^{[2
]}

Wang, Jin ^{[2
]}

Tian, Wei ^{[2
]}

Dickson, Keddy Wornyo ^{[2
]}

机构：

[1] Nanjing Univ Informat Sci & Technol, Jiangsu Engn Ctr Network Monitoring, Nanjing, Jiangsu, Peoples R China

[2] Nanjing Univ Informat Sci & Technol, Sch Comp & Software, Nanjing, Jiangsu, Peoples R China

来源：

INTERNATIONAL JOURNAL OF GRID AND DISTRIBUTED COMPUTING | 2013年 / 6卷 / 01期

基金：

美国国家科学基金会;

关键词：

Small Files; transfer performance; GridFTP; Compression; Meteorological Grid;

D O I：

暂无

中图分类号：

TP31 [计算机软件];

学科分类号：

081202 ; 0835 ;

摘要：

This paper focuses on performance enhancement of large scale of small files transfer, which is critical to the performance of meteorological grid. GridFTP and compression techniques are used to optimize the efficiency. The transfer parameters are configured before transmission, such as extended block mode (Mode E), TCP buffer size. Compression is used to compress files into single file, in order to eliminate negotiating time before transmission. During the course of our research, we sacrifice the disk space for avoiding the vicious recurrence of compression-decompression-compression. In the end, the experiments show the advantages of our approach.

引用

页码：25 / 33

页数：9