Using Pattern-Models to Guide SSD Deployment for Big Data in HPC Systems