Troubleshooting errors

This page describes error messages you might encounter when querying data with Ads Data Hub, and provides troubleshooting guidance.

Error messages can be found in the following locations:

While editing a query in the Ads Data Hub UI, error messages will appear in the validation box above the query text. You can expand the error text by clicking on Show validation.
After running a query that fails to complete, the Details field specific to the job will contain an error message in the Error field. A list of your recent jobs can be found by clicking on Jobs in the Ads Data Hub UI.
You can retrieve error messages using the Ads Data Hub API, using the operations resource.

This table does not include all possible error messages. If you don't see the relevant error here, or the suggested steps don't resolve the issue, contact support.

Error message	Description	Troubleshooting
Resources exceeded during query execution: The query couldn't be executed in the allotted memory.	This error is returned when your query requires too much memory to execute.	Avoid reaggregation. Use a smaller data date range. Break up your query into multiple, smaller queries. Then, combine the outputs in BigQuery. Alternatively, try using temporary tables to split the query up. Ensure that the larger table is on the left side of any `JOIN`. Review Ads Data Hub's best practices. Review additional troubleshooting in the BigQuery error messages article.
BigQuery error. Error code: Internal	This error is returned for various reasons. Common causes of this error include:
	A temporary issue or a timeout	Try running your query again.
	When using `OFFSET()` or `ORDINAL()`, if the index is out of range	Remove `OFFSET()` or `ORDINAL()`, or rewrite, using `SAFE_OFFSET()` or `SAFE_ORDINAL()` instead. ( ref.)
	When there is a data overflow caused by re-aggregations.	Avoid re-aggregation in query. Avoid joining unaggregated to aggregated and then aggregating once more.
	When the result has an array which contains a `NULL` element	Review BigQuery rules concerning `NULL` elements in arrays.
	If none of the these scenarios apply to your query, reach out to Ads Data Hub Support to diagnose the error.
Not found: Dataset `<myproject:mydataset>` was not found in location `<regioncode>`.	This error is returned when you attempt to output query results to a BigQuery dataset that is not in the same region as your Ads Data Hub instance.	Re-execute the job, pointing to a dataset that is in the same region as your Ads Data Hub instance. Copy or recreate the data in a dataset that is in the same region as your Ads Data Hub instance. For Ads Data Hub instances in the US or EU, specify the multi-region location `US` or `EU`; specifying a specific region, such as `us-east1` or `europe-west1` will cause jobs to fail.
	Ads Data Hub Support cannot modify or change the region assigned to your Ads Data Hub instance.
Table cannot be accessed in BigQuery.	This error returns during results preview, if you don't have access to the output table in BigQuery.	Have someone in your organization with the proper credentials in BigQuery grant you permission to view the table. Re-execute the query, specifying an output table for which you have permission to view in BigQuery.
Your query didn't pass Ads Data Hub privacy checks. Make sure your query meets aggregation requirements and difference checks. Error code: 0400	This error returns when query results are insufficiently aggregated or they are too similar to previous results.	Ensure that each result row meets the minimum required user count (50 users in most cases; 10 users when only clicks and conversions are accessed). Use sandbox data when testing queries. Review Ads Data Hub's best practices.
System error. Please contact support. Error code: Internal	This error is returned for various reasons. Common causes of this error include:
	Querying a first-party table that is not in same region as your Ads Data Hub instance	Copy first-party data to a dataset that is in the same region as your Ads Data Hub instance.
	Timestamp overflow	Use `TIMESTAMP_MICROS()` to convert time fields (e.g., `query_id.time_usec`) from microseconds to a timestamp.
	Querying data hosted in Google Drive storage	Host the data in Google Cloud storage.
	Billing has not been enabled for the project	Add a billing account for the project in Google Cloud Console. (Learn more.) Should you have any questions about that task contact the BigQuery support team or see this BigQuery Help Center article for more assistance with billing issues.
	If none of the these scenarios apply to your query, reach out to Ads Data Hub Support to diagnose the error.
The query contains `<number>` user sets, which exceeds the 100000 limit	This error is returned when the query results would contain more than the limit of 100,000 user sets. A user set is the most basic unit of protection in Ads Data Hub. A query gets one user set each time it retrieves data from a user-level Ads Data Hub table.	Reduce the complexity of the query. Reduce the number of `JOIN`s in the query, particularly for user-level Ads Data Hub tables. Try querying over a shorter date range. Break up the query into multiple, smaller queries.
Reminder: Queries that use 1PD tables require that you update your consent settings.	This error is returned when a query references first-party data (1PD) tables, if you have not acknowledged that you have obtained user consent.	Obtain and acknowledge user consent. Remove the references to the first-party data tables.
Tables from separate Google services must not be combined. Aggregate each Google service individually and combine after	This error is returned when a query combines data from different Google services, such as using a disallowed combination of policy-isolated tables.	Separate each Google service into a single result, either by Limitng each query to a single Google service Using `CREATE TABLE` statements with `OPTIONS(privacy_checked_export=true)`.
Noise injection is disabled due to query validation failure. Unexported tables must be accessed with the same privacy mode that they were created: `<tablename>`.	When a query reads from an unexported table, you must use the same privacy mode that was used to create the unexported table: both using difference checks or both using noise.	Run this query in a different privacy mode. Create the unexported table in the chosen privacy mode and then rerun this query in the same privacy mode. Create the unexported table in the same query that reads from the table. Export the table, instead. See Export multiple tables using a single query.
Setup failed. We've encountered issues setting up your account for UPDM. Please contact support via the buton below.	You might encounter this error if Domain Restricted Sharing (DRS) is enabled on your Google Cloud project.	Temporarily disable DRS on your Google Cloud project to complete UPDM setup. Once setup is complete, you can re-enable DRS. If you cannot disable DRS due to organizational policies, reach out to our support team for assistance.