Download presentation
Presentation is loading. Please wait.
Published byPercival Hicks Modified over 6 years ago
1
ENHANCING PERFORMANCE OF DATA MIGRATION VIA PARALLEL DATA COMPRESSION
2
OBJECTIVES Introduction. Related Work :- BBFTP.
Studying Different Data Compression Techniques. Application. Conclusions.
3
INTRODUCTION Data Migration. Data Compression. Overview Of Topic.
4
DATA MIGRATION Data migration is the process of transferring data between storage types, formats, or computer systems. It is required when organizations or individuals change computer systems or upgrade to new systems, or when systems merge.
5
DATA COMPRESSION Compression Means The Reduction In Data Size.
It Helps To Reduce The Consumption Of Resources.
6
OVERVIEW OF TOPIC What Is Today’s Need? What We Have? What We Can Do?
7
RELATED WORK :- BBFTP BBFTP : It is file transfer protocol for the files which are larger than 2 GB. Compression is introduced in it, mainly focusing on HTTP. FTP normaly not supports parallel I/O transfer. So this work can be added to BBFTP to support parallel servers .
8
STUDYING DIFFERENT TECHNIQUES OF DATA COMPRESSION
9
STUDYING DIFFERENT TECHNIQUES OF DATA COMPRESSION
TECHNIQUES OF COMPRESSION CLIENT-SIDE COMPRESSION (CC) . SERVER-SIDE COMPRESSION (SC) . SERVER-SIDE COMPRESSION (SC-2).
10
CLIENT-SIDE COMPRESSION(CC).
Each client compresses each of its output parts and sends them to a server. I/O servers receive compressed parts from clients and store them to disk. High Degree Of Parallelism. Needs Same Distribution Of Arrays In Memory And On Disk.
11
SERVER-SIDE COMPRESSION(SC).
Servers receive output data from clients during an I/O phase, compress them, and store them to disk. When server stores the data, it forces clients to wait. So I/O time increases. Performs Worst Than Client Side Compression(CC).
12
SERVER-SIDE COMPRESSION(SC2).
SC2 reads and compresses the stored output, and then migrates it. Only suitable for dedicated I/O servers. Less degree of parallelism during compression.
13
STUDYING DIFFERENT TECHNIQUES OF DATA COMPRESSION
14
Which Technique Should We Go For?
Let us consider some data sets. SCAR-B(Data set which includes physical and chemical components of earth’s surface). AVHRR(Data sets stores reflectance of earth). ZEUS-MP(Hydrodynamic Codes Used in astrophysics). And Apply All Three Techniques on Them. (cont..)
15
APPLICATION OF COMPRESSION ON DATA SETS
16
APPLICATIONS Predicting Compression Performance. Delta Compression.
How to predict? Sampling. Delta Compression. Consecutive difference between two arrays. Migration is faster than whole data migration
17
CONCLUSION we have seen three different approaches to enhance or improve that migration performance using parallel processors And we observed that CC gives the best performance overall, and it will be likely to dominate the NC approach either in a single migration or a real simulation run in a current typical parallel environment, with reasonable processing power. CC has no significant drawbacks when we do not have to reorganize the data.
18
REFERENCES Jonghyun Lee, Marianne Winslett, Xiaosong Ma, Shengke Yu, “Enhancing Data Migration Performance via Parallel Data Compression”,.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.