A Variable-Grain Consistency Maintenance Scheme for Shared Data on Emergency and Rescue Applications

• Main Authors: Jia-HaoYang, 楊家豪
• Other Authors: Ce-Kuen Shieh
• Format: Others
• Language: en_US
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/47066063941503491658

碩士 === 國立成功大學 === 電腦與通信工程研究所 === 98 === In the scenario of the emergency and rescue operations, information sharing is the most important factor that affects the success and failure of the entire operation. However, efficient information sharing is difficult to achieve in such a scenario because there is no communication infrastructure existed at the disaster sites. Generally Speaking, the network condition is relatively reliable in the intra-site environment and relatively unreliable in the inter-site environment. Moreover, the network partitioning problem may occur between two sites. Therefore, the replication technique in data grid should be used in emergency and rescue application for improving the efficiency of the information sharing. However, the data consistency problem occurs between replicas. In this context, we propose a middleware called Seagull that can transparently manage the data consistency, designated as Seagull for the situation of the emergency and rescue applications. Seagull adopts the optimistic replication technique that provides high availability, and adopts the pessimistic replication technique in the intra-site environment that provides strong consistency. Moreover, it adopts an adaptive consistency granularity strategy that achieves the better performance of the consistency management because this strategy provides higher parallelism when the false sharing happens. Lastly, Seagull adopts the transparency data consistency management scheme, and thus the users do not need to modify their source codes to run on the Seagull.