Font Size: a A A

Key Technology Of Large-scale Database Disaster Recovery System To Eliminate Duplicate Data

Posted on:2012-03-16Degree:MasterType:Thesis
Country:ChinaCandidate:C J WangFull Text:PDF
GTID:2208330332986728Subject:Software engineering
Abstract/Summary:PDF Full Text Request
As the economic globalization's deeply developing and rising of the information industry, international enterprises is emerging. The enterprises are processing millions of explosive production data from tens of millions of users. The data transfer and data storage capacity of data center are particularly important. Statistics show that there are many similar, even the same, data (called redundant data) in these new mass of data. Therefore, how to eliminate redundant information, save bandwidth and save space of storage, so as to reduce the burden of the data center, has become the key problem urgently to be solved. Therefore, this thesis proposes Deduplication technology.Based on the intensive study of deduplication technology, this thesis completed the following work.Firstly, an innovative Encryption Scheme for deduplication System has been proposed. This thesis described ESDS from three participants of the encryption, and then analyzed the Security of system when encountering a variety of attacking. ESDS is able to provide reliability data protection when users have needs for transferring encrypted data.Secondly, this thesis designed a large structure and function Disaster recovery system. The system design based on comprehensive use of three mechanisms, which are preliminary accessing, data storage based on data locality, metadata access based on inertial property-write with buffer. So as to improve the efficiency of deduplication;Finally, in the Part of test, this thesis designed a method for constructing test data set originally, which is used for the test of deduplication module. This thesis designed several experiments, which contain a couple of comparison for one function of different methods of deduplication. And then drew the conclusion by analyzing the test data.The results show that the deduplication module designed in this thesis is able to run in two different chunking ways, which contain fixed length chunking and Content-Defined Chunking, and had a fast backup speed.
Keywords/Search Tags:disaster backup, chunking, encryption, ESDS, deduplication
PDF Full Text Request
Related items