摘要
为了减少序列化产生的数据量,以基于网格的PDE.Mart为实验平台,使用APACHE AXIS,研究AXIS的不同服务类型,比较压缩前后直接SOAP序列化、Base64序列化以及字节流序列化后的网络传输数据量,提出一种二进制序列化与压缩相结合的方案,减少生成的消息数据,从而提高数据传输性能,使网格计算平台能适应较大规模数据的求解。
To decrease the data generated by serialization, this paper uses grid-based PDE.Mart as platform, investigates different service types of Apache AXIS, compares the amount of data generated by different serialization protocols like SOAP serialization, Base64 serialization and byte stream serialization before and after compression. A hybrid approach that combines byte stream serialization and compression technology is proposed, which can significantly decrease the data being transferred in the grid, so as to improve performance of data transfer and enable grid computing system to solve large scale problems.
出处
《计算机工程》
CAS
CSCD
北大核心
2008年第16期89-91,共3页
Computer Engineering
基金
江苏省高校自然科学基础研究基金资助项目(07KJD520009)
关键词
序列化
压缩
网络服务
serialization
compression
Web service
