Our app is 100% U.S. based with out any double byte characters. Each column is an ascii datatype. There are no unicode datatypes used. I have read that to prevent the data from being converted from unicode to ascii by the jdbc driver, a subprotocal driver named "inetdae7a" should be used. I am looking into the accuracy of that statement. Does anyone have experience in this type of issue.
when I use the new subprotocal inetdae7a, which doesn't use a unicode and therefore no conversion occurs, table access improves by 50% but when accessing by views there is negligable improvement. The table and view are exact copies of each other for a different test that is going on at the same time. My assumption was that the improvement would be equal or greater for the view access than the table access. Anyone have any ideas as to why the table time improved greatly but the view times remained essentially constant?