Optimization of application execution in the ViroLab Virtual Laboratory

The objective of the presented work is to describe an optimization engine for the ViroLab Virtual Laboratory runtime.
The Laboratory specific model - invocation of operations on special objects which reside on Grid resources - imposes a new approach to optimization of Grid application execution. Such an approach is presented in the shape of the GridSpace Application Optimizer (GrAppO) - a module responsible for optimization of exectution of the applications developed in the ViroLab Virtual Laboratory - supported by monitoring and provenance systems. We describe the architecture of the optimizer, and report on the results of tests which demonstrate the need for the monitoring system.