Abstract :
Volunteer computing systems such as BOINC use several interacting scheduling policies, which must address multiple requirements across a large space of usage scenarios. In developing BOINC, we need to design and optimize these policies without direct access to the target nodes (volunteered PCs). To do this, we developed an emulation-based system that predicts the policies´ behavior in specific scenarios. This system has been useful in the design and evaluation of policies, in software development, and in the resolution of problems occurring in the field.