I am running QODBC as a mechanism to get data from QuickBooks Enterprise for a report processor I have built.  Extracting  data from the client table for exmaple is slow (but tolerable  - with only 50 client records) HOWEVER, the moment that you apply any kind of sort becomes EXTREAMLY slow.  I have worked with many databases, most of which have index tables for sort optimization.  From within Quickbooks, sorting (even on client tables of hundred of records) happens pretty quickly.  Why is OODBC so slow. Based on my estimates to sort by client name on a table with 300-500 could take up to 2 -5 minutes. That will not be acceptable to end users.

Any ideas.  I should be able to read the Sybase database behind it directly with MUCH better results, I hope.

Thanks

Rick

Most likely you are doing something like:

SELECT * from Customer order by CompanyName

which will run VERY slowly as QODBC has to sort the result and display the correct current open balance for each customer, row by row, while this will run much faster:

SELECT "Name","LastName","FirstName", "CompanyName",
"BillAddressAddr1", "BillAddressAddr2","BillAddressCity",
"BillAddressState","BillAddressPostalcode"
from Customer order by CompanyName

and can be refreshed even faster by using:

SELECT "Name","LastName","FirstName", "CompanyName",
"BillAddressAddr1", "BillAddressAddr2","BillAddressCity",
"BillAddressState","BillAddressPostalcode"
from Customer NOSYNC order by CompanyName

See: How do I setup the QODBC Optimizer? Where are the Optimizer options?  for more info.

The stored procedure sp_columns actual shows the jump-ins that can be used in the WHERE clause to locate a row instantly:

sp_columns customer

For example:

SELECT "Name","LastName","FirstName", "CompanyName",
"BillAddressAddr1", "BillAddressAddr2","BillAddressCity",
"BillAddressState","BillAddressPostalcode"
from Customer Where FullName='ABC Company'

uses the FullName jump-in. If you don't use a jump-in, QODBC will need to do a table scan to locate the rows to be returned.