Splunk DB Connect 2: How to create a DB Input when NO Results are retrieved from the SQL query in that instance?

[dsuryaavinash]

we observed a few issues with DB connect 2 app

In DB connect 2 , the “Continue” button is disabled and we are unable to create a “New DB Input” when no results are retrieved in that instance. When trying to create DB Inputs for queries that may fetch results later on, this is not possible. Can you please check and let us know an alternative to this?

[araitz · Accepted solution]

You would need to create the db input manually via inputs.conf. Because there is no data in the source table, the UI can't guide you through selecting timestamp and rising column fields in order to prevent configuration errors.

[tweaktubbie]

It’s still kind of weird Splunk mentions this is done to prevent configuration errors, but the only workaround now is to go to another input option which does NOT check on validation 😉

However, the workaround from dsuryaavinash (thnx!) to this suggestion is a valid one: create in the regular DBConnect V2 interface the query and settings and use a dummy query like ‘select 1 from DUAL’ and after saving edit it in the data inputs/DB Connect input service. (but again, this does not validate if the query is correct). Everyone I speak using DBConnect can't understand this logic.

[Skorfulose]

Hello Alex,

in the inputs.conf one needs to provide input_timestamp_column_number and tail_rising_column_number. How do we get the column number manually if there is no result to the query yet?
How are the columns counted anyway? And why is *column_name not sufficient?

Regards,
Thomas

[dsuryaavinash]

Ya , but in any case it is a scheduled DB input and if that instance doesnt produce an output , we will be unable to create the DB input. We worked on DB Connect 1 and i dont remember having faced this issue.

By the way we figured out a solution using -- DATA Inputs --> Splunk DB Connect Input Service --> Add new

Thanks for the reply 🙂