I am looking to run this algorithm on one million nodes (A1.....An) repeatedly doing route.path('A1'...'K') So only the starting node changes, whereas the ending node will stay the same. In spite of this, I would like to know whether any benchmarks have been performed regarding a large dataset?