A New Partitioning Algorithm to Enhance Performance of SOAP Message Security Processing

Abstract

Web Service Security (WS-Security) processing suffers from performance bottleneck. With large data centric web services and hardened schema it becomes even worst. A new partitioning algorithm with parallel stream based security processing model is developed for large Simple Object Access Protocol (SOAP) messages. The aim is to partition SOAP message and to distribute the processing load on multiple cores in order to enhance the performance of SOAP message security processing. There are several steps involved in this approach but SOAP message partitioning is mainly focused in this paper. Since the efficiency of SOAP security processing depends on both Extensible Markup Language (XML) data size and its structure, these factors are considered while developing partitioning algorithm. Each partitioned SOAP message data is uniformly distributed to parallel running instances of stream based security processor so that the load among cores can be balanced. It is observed from the results that the implementation of new partitioning algorithm with parallel stream based WS-Security processing model has significantly improved the performance of SOAP message security processing. The fundamental approach presented in this paper is useful for efficient security processing of large SOAP messages.