Traditional performance monitoring methods have an overhead that is proportional in both space and time to the spill rate. They are not scalable in a distributed computing environment of hundreds or even thousands of managed nodes; the large quantity of data collected cannot be pushed or pulled efficiently across the network for displaying/reporting. In this paper we propose a few scalable alte...

In this paper we give a critical overview of recent development in MIMO control performance monitoring. We discuss a number of MIMO control benchmarks including minimum variance, LQG, and user selected benchmarks. Performance measures are extended from variance based measures in SISO control to covariance based measures in MIMO control. Pros and cons of various benchmarks are discussed. The dia...

Dependable performance measurement is a common requirement for all on-line services. The ongoing tendency to outsource not only infrastructure, but also software parts to several suppliers, creates a new challenge to providers. Traditional solutions cannot easily monitor service performance as perceived by end users, since providers do not have control over all hardware/software involved. Our r...

Virtualizers on Type II VMMs have been popular among non-commercial users due to their easier installation and use along with lower cost than those on Type I VMMs. However, the overall performance of these virtualizers was mostly found worse than the latter from VM user’s point of view and, the reasons remain yet to be fully investigated. In this report, we present a quantitative study of VMM p...

Due to the complexity of modern hardware, computer programs often cause complicated interactions when scheduled together. Users therefore frequently turn to specialized tools such as Performance Monitoring Units (PMUs) for diagnosing system issues. However, PMUs themselves are complex, so they require domain knowledge and a good understanding of the underlying hardware to use. Such obstacles pr...