Left join on two Spark dataframes results in a dataframe longer than the right one

Question

I have two Spark Dataframes:

• A: 100 records, columns: column-abc, column-bcd, column-cde, index1, index2
• B: 357888 records, columns: