Iterative GraphFrames AggregateMessages hitting memory limits

Question

I\'m using GraphFrame\'s aggregateMessages capability to build a custom clustering algorithm. I tested this algorithm on a small sample dataset (~100 items) and verified tha