We're extracting data from IBM Db2 data to load it in a data warehouse. We've found that the JDBC driver outperforms by far the ODBC/OLEDB iSeries drivers. For example, extracting one million rows takes seconds using the JDBC driver and three minutes using the ODBC driver. The only problem is that we can't use the JDBC driver. What can explain the huge difference in performance?
------------------------------
Teo Lachev
------------------------------