An empirical evaluation of OS endsystem support for real-time CORBA object request brokers

• Published in: Proceedings of SPIE, the international society for optical engineering Vol. SPIE-3969; pp. 113 - 128
• Main Authors: LEVINE, D. L, FLORES-GAITAN, S, SCHMIDT, D. C
• Format: Conference Proceeding Journal Article
• Language: English
• Published: Bellingham WA SPIE 2000
• Subjects: Applied sciences; Computer science; control theory; systems; Computer systems performance. Reliability; Distributed computer systems; Electronics; Exact sciences and technology; Hardware; Software; Telecommunications; Telecommunications and information theory; Teleprocessing networks. Isdn; Performance evaluation; Multimedia; Operating time; CORBA software bus; Platform; Scheduling; Windows; Real time system; Computer hardware; Object oriented; Measurement method; Service quality; Operating system
• ISBN: 0819435872; 9780819435873
• ISSN: 0277-786X

There is increasing demand to extend Object Request Broker (ORB) middleware to support applications with stringent quality of service (QoS) requirements. In particular, distributed multimedia applications require efficient and predictable performance from the middleware layer and below to deliver necessary end-to-end QoS. Although ORB middleware, such as CORBA, COM+, and Java RMI, offers transparent distribution, the QoS that it can provide depends largely on the support from the underlying endsystems and networks. This paper provides two contributions to the study of OS endsystem support for real-time ORB middleware. First, we empirically compare and evaluate the suitability of real-time operating systems, VxWorks and LynxOS, and general-purpose operating systems with real-time scheduling classes, Windows NT, Solaris, and Linux, for real-time ORB middleware. While holding the hardware and ORB constant, we systematically vary the OS and measure key platform-specific variations in latency, jitter, operation throughput, and CPU processing overhead. Second, we describe specific areas where these operating systems must improve so that ORB middleware will be predictable, efficient, and scalable enough to support the QoS requirements of multimedia applications. Our findings illustrate that general-purpose operating systems are not yet suited to meet multimedia applications with stringent QoS requirements. Our results underscore the need for a measurement-driven methodology to (1) provide an empirical basis for metrics that are crucial to multimedia computing and (2) to use these metrics to pinpoint and alleviate sources of priority inversion and non-determinism in real-time ORB endsystems.