I'm working with an Excel file of ~50,000 rows and 18 columns, connected via Custom SQL as described in http://community.tableau.com/thread/104070 for NULL-padding of entries that lack data. Loading metadata when making the connection takes four hours before the query itself begins to run, and incremental refreshing of the extract takes about as long. From the size of the datasets that I see mentioned in other posts here this seems rather long, even considering the extra work caused by the join. What I'd like to know is: are these ordinary times for this sort of task, and if not, where could the problem be and what can be done to fix it?
As an extra datapoint, when working with a subset of about 5000 rows the metadata stage took about 10 minutes, and the query a few more past that.
Robert, there may be other options depending on your exact situation.
Can you provide an example table workbook and Excel file that represents your situation, and the results your desire from this sample data?