Page 1 of 1

TI in 64-bit TM1 server cannot point to 32-bit ODBC Admin

Posted: Thu Feb 10, 2011 3:56 am
by harrytm1
Hi all,

I have searched through the forum on this topic. Currently, I'm using Windows 7 and the TM1 server is 64-bit.

I'm trying to develop a TI to connect to an Access DB. When I clicked the "Browse" button in the TI next to Data Source Name field, the list did not include the Access data source that I had created in the 32-bit ODBC Administrator panel.

If I try to set up the same Access data source using the 64-bit ODBC Admin panel, I'm only presented with two drivers: SQL Server and SQL Server Native Client 10.0.

Is there a way to get TI to point to the 32-bit ODBC Admin since I can't define the data source in the 64-bit ODBC Admin?

Thanks!

harry

Re: TI in 64-bit TM1 server cannot point to 32-bit ODBC Admi

Posted: Thu Feb 10, 2011 4:00 am
by Alan Kirk
harrytm1 wrote:Hi all,

I have searched through the forum on this topic.
Does this include the thread under the heading "64 Bit ODBC Driver Issues" in the FAQ? It would seem to be quite relevant.

Re: TI in 64-bit TM1 server cannot point to 32-bit ODBC Admi

Posted: Thu Feb 10, 2011 4:14 am
by harrytm1
Yes I did before posting. I posted this because that post did not offer a solution.

Re: TI in 64-bit TM1 server cannot point to 32-bit ODBC Admi

Posted: Thu Feb 10, 2011 5:17 am
by lotsaram
harrytm1 wrote:Yes I did before posting. I posted this because that post did not offer a solution.
Harry, that's because there isn't one. MS Access is a 32 bit application and has (as far as I'm aware) no 64 bit driver. There is no direct route to get data directly from Access into a 64 bit TM1 server, but there are any number of common workarounds, some of which are ...
- migrate the access application to a RDBMS that supports 64 bit ODBC drivers (such as SQL Server)
- if your TM1 application is small use 32 bit TM1 server
- use a 2nd 32 bit TM1 server as a data staging area
- rather than ODBC use flat files
- automated flat files: use a database tool that supports command line (such as querytool) to pass in an SQL query and generate a CSV file (triggered by ExecuteCommand in TI), then pick up the file

The first and last options are generally my preferred approach.