The ultimate objective of the proposed project is to provide the most accurate tools to run under real loads in order to monitor and trace sophisticated distributed (possibly real-time) systems. These tools will be used for system health monitoring, and performance tuning and debugging during development, operation and maintenance. To achieve this, work is required at three levels: i) new techniques and algorithms must be developed to solve a number of research issues; ii) the proposed techniques and algorithms must be integrated into the development and operation tool chain; iii) the improved tool chain must be tested under realistic conditions. As a consequence, the project relies on the close collaboration of several groups: the researchers specialized in the different fields relevant to the research issues, the tools development groups, and the industrial and governmental
users of these tools.
The research group is comprised of 5 professors, principal investigator and co-applicants, and 10 graduate students supported by a research associate, as detailed in the budget justication. The tools development groups include the Linux kernel developers (the LTTng project supervised by professor Michel Dagenais being a regular contributor), The Eclipse foundation (two Ericsson Canada and several IBM Canada employees are full time contributors), and internal groups at Ericsson, IBM, MontaVista, Nokia, Oracle, Red Hat, Wind River and others. Through LTTng, Ericsson Canada and Defence R&D Canada, the project will have a direct access to Linux kernel and Eclipse developers, to internal tools groups at Ericsson, and to industrial and governmental users of these tools.