I have a 1M and 10M rows tables (10 columns) I am uploading to a postgres DB using pandas.to_sql
.
I noticed that using method="multi"
severely reduces performance, 5 to 6x with chuncksizes 5k, 10k, 20k. even reducing chunksize to 100 didnt help.
Is this normal with postgres? 10 columns is not a lot for the performance to be that affected.