摘要: From the inception of digital storage, ensuring that data is not lost due to user error, malicious acts, and hardware failure has always been, still remains, a challenging open problem. This problem exacerbated by exponential increase in storage capacity, proliferation new media, our growing reliance on storage. Today, typical stores financial medical records, music movie libraries, photo albums, etc, loss some which can be catastrophic. The advent large robust networks made it possible replicate remote hosts protect from loss. Unfortunately, growth network bandwidth far outstripped both capacity ability fill it. Thus, most replication systems uniformly all are incapable protecting ever increasing amount data. One important observation created equal. Data such as commercial libraries be, given time, rebuilt. personal, health, much more difficult reconstruct. Since resources limited, they should used In this thesis we propose Policy Driven Replication (PDR) system prioritizes according user-defined policies specify what protected, failures, extent. By prioritizing replicated, conserves limited protects high-priority high-probability failures. PDR userlevel process hooks into file system. It notified creation modification events, replicates specified file's policy. In addition, replica nodes policy monitored for liveliness ensure followed. provides model describe generic plug-in interface facilitates appropriate interfaces manage translate these set nodes. Replica node selection sensitive topology so hotspots message storms created.