Is it possible to join and get common Id%27s from two different Teradata Servers tables in SAS EG%3F

[msuri113]

I am unable to get common Id's from two different Teradata server tables.

[Quentin · Accepted solution]

Please show the code you have tried, and describe the problem.  Are you getting errors in your log, or just unexpected results?  If the results are unexpected, please describe what you are getting for results, and what you expect.

Assuming you have SAS libraries pointing to the two different servers, it should be possible to join them.  But I assume you will need to pull the data into SAS first, unless there is a way in Teradata for the two Teradata servers to see each other's data.  If the teradata servers can see each other's data, then from SAS you should be able to send an explicit pass-through query to one of the teradata servers and have it run the join and then send only the results of the join back to SAS.

I haven't played with Teradata much.  But I know with SQL server you can configure two different SQL server instances on different servers to see each other's data.

[msuri113]

Thanks for quick reply Quentin.
We need common Id's from both teradata servers tables based on where condition We filtered two server tables.We have big data in both Teradata servers, so, the reason if I filter the data from Teradata server-1 it will get huge Id's and server-2 as well. My question is I want to execute both Teradata server tables with join condition and apply the where clause same time then will get common Id's .

[PaigeMiller]

So @Quentin asked "Please show the code you have tried, and describe the problem. Are you getting errors in your log, or just unexpected results? If the results are unexpected, please describe what you are getting for results, and what you expect." You didn't do that.

[Quentin]

If you were submitting native teradata queries on one of the teradata servers, would you be able to query data from the second teradata server?

If not, then I don't see how you can avoid pulling the ID's from both servers down to SAS to do the join.

I suppose you could try pulling just one set of ID's from teradata  server A and then you could push the ID's up to teradata server B and do the join in teradata.  So that would only require transferring one set of ID's, but you would have to pull it from server A and then push it to server B, so it's still a lot of time moving data.

I would work with your admins to see if you can get a fast connection between the two teradata servers.