Cathi,

>You used the SEEK function here. I am using remote views for all data access with the back end being SQL Server. Would I be better off using the SEEK or the LOCATE function? These tables could get quite large in the future. I will almost always SEEK/LOCATE on the primary id. To use the SEEK function I assume I would need to create a temporary index when the view is created. I would think this could take a long time.>

As I'm sure you know, your goal for optimization purposes is to return as few records as possible from the SQL Server database. You can do this by using a parameterized remote view or using SQL pass-through.

If you are certain that you have returned a small number of records it may be faster to use LOCATE rather than build a temporary index then issuing a SEEK...even though LOCATE needs an index to be Rushmore-optimizable. If you use LOCATE without an index, you'll get better performance if you include the NOOPTIMIZE clause which turns off Rushmore optimization.

However, if you have returned a large number of records it would probably be faster to build the index before SEEKing, especially if you are doing multiple SEEKs on the returned set of records.

Regards,