JDBC Source getting "Identity attribute [XXXX] was not found" at aggregation

[jsosa (Julian Sosa)]

Hi, I am getting following error when aggregating a JDBC source. Is weird because in my query, I explicitly ensure that this attribute is not null (… AND E_MAIL IS NOT NULL).

Also, if I start a new aggregation, it adds more accounts (about 1500 accounts each time), but is far from getting all accounts (about 45k):

In my schema, E_MAIL is both account id and name, and is the attribute used for correlation.

[DFuyal (Danish Phuyol)]

Maybe use firstValid transform in email field. Fetch the email as account attribute from the source in transform and if it is null, set something static in transform. eg; “NONE” so that your email field is never null. Check if it works.

[Swegmann (Sebastian Wegmann)]

Can you share the query used?

Does the same query work without errors in your SQL manager?

We have had similar errors in the past due to order problems, although different error message. We fixed those errors simply by adding an order clause at the end of our query.

Try:

ORDER BY [E_MAIL]

You can also try:

ORDER BY [E_MAIL] COLLATE Latin1_General_CI_AS

If trying different order clauses doesn’t solve your problem, there’s probably something you’re missing. I recommend trying to troubleshot your issue in your SQL Manager for the database.

Are you sure there are no instances of null values for email? Check in your SQL manager with:

SELECT *
FROM your_table_name
WHERE [E_MAIL] IS NULL;

Or this:

SELECT *
FROM your_table_name
WHERE ISNULL([E_MAIL], '') = '';

[jsosa (Julian Sosa)]

Hi, sorry to forgot to close this post and thanks for the help!

Cause was that client had several garbage values in this attribute, so excluding these rows aggregations worked as espected.