Error SQLExecute: [Simba][BigQuery] (115) Operation timeout. Timed out pol

[Heriberto_Prieto]

Hello everyone,

Trying to pull data from Google big query using the Simba ODBC driver but every time a query takes more then five minutes it timeouts. Running the query directly on google console the timeout doesn't happen. Does anyone know if there is a way to increase that timeout limit?

[apathetichell · Accepted solution]

This is controlled by the ODBC driver - edit it in ODBC config for your driver:

https://storage.googleapis.com/simba-bq-release/odbc/Simba%20Google%20BigQuery%20ODBC%20Connector%20...

[Heriberto_Prieto]

Thanks,

going through the documentation found out that adding a timeout to the connection string solved the issue

[JMB001]

@Heriberto_Prieto Could you please share details on how you did this with the ODBC - I don't see an option in Advanced Options in the DSN Setup. 

@STP - this timeout message is happening on Gallery runs.

[apathetichell]

Are you connecting via JWT on Gallery/Server? page 100 of the guide I posted shows you the setting you can control for timeout.

[Heriberto_Prieto]

Hi JMB001,

If you are using windows the timeout has to be added to the connection string, the driver won't have that setting, here is a screenshot on how to add the timeout to the connection

[JMB001]

Thanks guys.

I wasn't clear enough: I get this issue when OUTPUTTING to BigQuery using a InDB Connection File to open the connection in the Connect In-DB tool in workflows that are entirely In-DB tooling.

I also get the error using the InDB Connection File and output using the DCM - Google Bulk Loader.

I can run them locally but not on our Gallery Server. But locally, I have no special line/code/setting for timeout.

@apathetichell - I've looped in our Server admin as I dont know how to answer that.

@Heriberto_Prieto - thanks for the screen shot. Where is that connection string location?

[apathetichell]

If I was your architect - I would recommend doing this all via scripted GCP CLI where your Server has preauthenticated to a specific GCP project/account - or connects in a wofklow. I would recommend moving your data to a storage bucket and having BQ query the storage bucket  vs uploading directly to BQ.