PDMRTS: Multiprocessor Real-Time Scheduling Considering Process Distribution in Data Stream Management System

• Published in: Digital Information and Communication Technology and Its Applications pp. 166 - 179
• Main Authors: Alemi, Mehdi, Safaei, Ali A., Hagjhoo, Mostafa S., Abdi, Fatemeh
• Format: Book Chapter
• Language: English
• Published: Berlin, Heidelberg Springer Berlin Heidelberg
• Series: Communications in Computer and Information Science
• Subjects: Data Stream; DSMS; Multiprocessor; Partitioning; Real-Time
• ISBN: 3642220266; 9783642220265
• ISSN: 1865-0929; 1865-0937
• DOI: 10.1007/978-3-642-22027-2_15

In Data Stream Management Systems (DSMSs), as long as continuous streams of data are arriving in the system, queries are executing on these input data. Regarding high volume of input data, having high processing capacity by using multiple processors is non-negligible. Also, many applications of DSMSs, such as traffic control systems, and health monitoring, have real-time nature. To support these features, this paper aims at developing an efficient multiprocessor real-time DSMS. To achieve efficiency, a multiprocessor real-time scheduling algorithm is proposed based on partitioning approach. In this algorithm, each received query has a chance to fit into any processor with first fit assignment. If it could not fit due to its utilization then that query is broken into some queries with smaller processing capacity based on utilization of processors. We conduct performance studies with real workloads. The experimental results show that the proposed algorithm outperforms the simple partitioning algorithm.