Timelines generated with timeline.py script

k-means iterations

Blank job

Large dataset (data pre-processing, 350GB) scale-out

Small dataset (k-means iteration, 2GB) scale-out

Large dataset (pre-processing, 350GB) overheads