The same graphs with different sizes

Question

I'm building big pan-genome graphs in Java Embedded mode. Considering a fixed topology, when I build it in one round the size of database is a bit less than when I build a sub-graph of it and add the rest of graph to it later in another execution of the application. What would be the possible reasons?

Answer 1

Your question is rather vague. I guess the store files itself will have the same size ( data/graph.db/*store.db ). The difference in size might be caused by the transaction log files ( data/graph.db/*transaction.db* ).