Clustering for new data in Role-stage-discovery
clustering for all data once. how should we process later data? cluster everything again or add nodes to existing clusters.
how should we fetch new data? currently just clearing the db and fetching all again.