The objective of this track is to predict and monitor the impact of activating sources of tracing data and reactive measures. To this end, it is necessary to extend the system model to complex heterogeneous networked systems, adding to the multi-level systems (multi-core system, virtual machines, system libraries and applications) the network as well as resources such as the memory (capacity and bandwidth) and the I/O devices. This extension to the original project allows the proposed Tracing and Monitoring algorithms and techniques to be applicable and tested in large complex heterogeneous production systems, and therefore help develop algorithms better suited for scaling to extremely large systems.
A first part is to have an extensible state model with associated state history database. This allows storing and accessing efficiently the extended model used for impact prediction.
The second part is to have modules which track the system resource usage, characterize the impact of activated tracing probes, and can predict the impact of activating additional probes on the different resources.