Optimization of request processing times for a heterogeneous data aggregation platform

The GRADLC heterogeneous data aggregation system allows for flexible expansion by connecting to new data storages, and provides researchers fast aggregated access to heterogeneous data of different astroparticle physics experiments, while reducing the storage load.
This flexibility comes with necessity to balance user requests in the queue with respect to various request processing times for the different storages, releated to different data processing policies and opportunities on a particular storage.

In order to handle this problem, a mathematical model for data aggregation was developed. Based on this model, a minimization problem for request processing times for the GRADLC system was formulated and investigated. Using simulation modeling a job shop scheduling algorithm was determined, which outperforms the straight forward FIFO approach in regards to the average request processing times.