الفهرس 
الغلافOnly 14 pages are availabe for public view
 |  | from | 188 |  |  |
| | | from | 188 | | |
Abstract
Partial Data Replication is proposed as a Disaster Recovery and Business Continuity (DRBC) Solution for Prevention, Detection and Recovery of critical services aim to increase availability, fault tolerance and system performance. The research introduces Partial Data Replication as a Pick-and-Run approach identifying the critical data to replicate during the peak hours, while the less critical replicates less frequently to avoid performance bottlenecks experienced by the existing approaches. PDR Manager is the proposed middleware to handle partial data replication across different sites by segregating data into groups replicating at different
frequencies. While most partial replication approaches focuses on partitioning physical data to
different physical sites, the proposed approach offers multi-dimensional vision based on replication timestamps, frequencies and business logic. Group Writeset Analyzer is the engine
proposed to monitor and analyze the database activities and optimizes the Partial Data
Replication Groups Writesets required for PDR technique. It gives an open mind approach for
temporarily compromising the data integrity at the DR site to ensure high availability and better
Point of Recovery in case of disaster. The survey analysis was done on the banking and financial
institutions as one of the business critical domains where failure/downtime factors highlights
”Data Loss” as the highest risk. The proposed PDR technique showed 50% decrease in the data
volumes