is proposed to solve the problems in grid data interoperation
such as heterogeneous storage resources
different job requirements
and numerous failures. The data-gateway is a professional solver for data interoperation that can make the data interoperation as an independent and general modular separated from job inter-operation
and promote parallel processing of the interoperation tasks. A job description language based on ClassAD is introduced in the data-gateway
which can provide a very flexible and extensible data model
and can be used to describe arbitrary services and constraints
as well as the scale mechanism in a plug-in style so that the scalability of the data-gateway for heterogeneous storage resources is improved well. Dynamic protocol selecting mechanism and failure recovering mechanism used in thedate-gateway can provide a nice environment like failure-free network and storage devices
and enhance the system's availability. Experiment results show that the data-gateway can implement data interoperation between heterogeneous grid infrastructures rapidly
recover failures automatically without any human interaction
and complete all kinds of data interoperation tasks efficiently.
关键词
Keywords
references
KOSAR T, LIVNY M. Stork: making data placement a first class citizen in the grid [C]∥Proceedings of 24th IEEE International Conference on Distributed Computing Systems. Piscataway, NJ, USA: IEEE, 2004: 342-349.
RAMAN R, LIVNY M, SOLOMON M. Matchmaking: distributed resource management for high throughput computing[C]∥Proceedings of the 7th IEEE International Symposium on High Performance Distributed Computing. Piscataway, NJ, USA: IEEE, 1998: 28-31.
YAN Bingheng, WU Zhongxin, YANG Dongbo, et al. Experiences with the EUChinaGrid project-implementing interoperation between gLite and GOS [C]∥The 2nd IEEE Asia-Pacific Service Computing Conference. Piscataway, NJ, USA: IEEE, 2007: 224-231.
SURRIDGE M, TAYLOR S, ROURE D D, et al. Experiences with GRIA:industrial applications on a web services grid [C]∥1st International Conference on e-Science and Grid Computing. Piscataway, NJ, USA: IEEE, 2005:98-105.