2-D tiles declustering method based on virtual devices

Generally, 2-D spatial data are divided as a series of tiles according to the plane grid. To satisfy the effect of vision, the tiles in the query window including the view point would be displayed quickly at the screen. Aiming at the performance difference of real storage devices, we propose a 2-D tiles declustering method based on virtual device. Firstly, we construct a group of virtual devices which have same storage performance and non-limited capacity, then distribute the tiles into M virtual devices according to the query window of 2-D tiles. Secondly, we equably map the tiles in M virtual devices into M equidistant intervals in [0, 1) using pseudo-random number generator. Finally, we devide [0, 1) into M intervals according to the tiles distribution percentage of every real storage device, and distribute the tiles in each interval in the corresponding real storage device. We have designed and realized a prototype GlobeSIGht, and give some related test results. The results show that the average response time of each tile in the query window including the view point using 2-D tiles declustering method based on virtual device is more efficient than using other methods.

[1]  Peter Widmayer,et al.  Distributing a search tree among a growing number of processors , 1994, SIGMOD '94.

[2]  Alex F. Bielajew,et al.  Fundamentals of the Monte Carlo method for neutral and charged particle transport , 2000 .

[3]  Zhou Xing A Scalable and Distributed Dynamic Interval Mapping Algorithm , 2006 .

[4]  Christian Scheideler,et al.  Efficient, distributed data placement strategies for storage area networks (extended abstract) , 2000, SPAA '00.

[5]  Noam Rinetzky,et al.  Towards an object store , 2003, 20th IEEE/11th NASA Goddard Conference on Mass Storage Systems and Technologies, 2003. (MSST 2003). Proceedings..

[6]  Gregory R. Ganger,et al.  Object-based storage , 2003, IEEE Commun. Mag..

[7]  Zheng Sheng Network Geographic Information System Architecture Based on Object-Based Storage , 2008 .

[8]  Compact , Adaptive Placement Schemes for Non-Uniform Capacities , 2002 .

[9]  Witold Litwin,et al.  LH*—a scalable, distributed data structure , 1996, TODS.

[10]  Liu Zhong A Data Object Placement Algorithm Based on Dynamic Interval Mapping , 2005 .

[11]  Robert Devine,et al.  Design and Implementation of DDH: A Distributed Dynamic Hashing Algorithm , 1993, FODO.

[12]  Ronald Fagin,et al.  Efficiently extendible mappings for balanced data distribution , 2005, Algorithmica.

[13]  Zhongmin Li,et al.  An object-based storage model for distributed remote sensing images , 2006, Geoinformatics.

[14]  Witold Litwin,et al.  LH*RS: a high-availability scalable distributed data structure using Reed Solomon Codes , 2000, SIGMOD '00.

[15]  Ethan L. Miller,et al.  A fast algorithm for online placement and reorganization of replicated data , 2003, Proceedings International Parallel and Distributed Processing Symposium.

[16]  Tore Risch,et al.  LH*G: A High-Availability Scalable Distributed Data Structure By Record Grouping , 2002, IEEE Trans. Knowl. Data Eng..