Performance analysis for real-time grid systems on COTS operating systems

Computer technology for communication has become an integral aspect of daily operation. The exponential growth of internet services with dynamic inquiries in such areas as manufacturing, business, air traffic control and mission critical systems demands that there be quick, reliable and safe use of services. Each service must contain QoS metrics to assure security, performance and accuracy. A new paradigm of resource management middleware techniques is in this paper presented which can provide QoS for dynamic, distributed real-time systems on Common Off The Shelf (COTS) operating systems. Accommodation of dynamic environments enables the middleware to carefully consider an efficient design of resource profiling, resource needs estimation, resource unification, and performance analysis (or compliant with schedulability analysis) infrastructure providing significant benefits for QoS management on COTS operating systems. First, the use of low-cost COTS systems is extended to real-time computing without changing the operating system. Further, experiments for response time analysis confirm that the worst-case analysis poorly utilizes computational resources. Finally, it is shown that the new method of middleware design employing scalability of software and hardware system can be easily applied to legacy systems to manage resources efficiently for quick, reliable services and accurate QoS.