@kapa.ai we have some tests failing because of a

abort query

, not because the test actually failed. The thing is, Elementary still treats this as a failed test which feels a bit confusing, because the test couldn't actually be ran, so the underlying data may very well be correct

Database error in test dbt_utils_unique_combination_of_columns_braze_user_canvas_behavior_entry_event_id__conversion_event_id (models/core/braze/braze_schema.yml)
  abort query
  compiled code at target/run/dbt_analytics/models/core/braze/braze_schema.yml/dbt_utils_unique_combination_o_f77e051403fb31c21…

This is a valid concern. Based on the Elementary documentation, there's an important distinction between test outcomes: • An error means dbt could not run the SQL at all (e.g., syntax mistake, missing table, broken macro, or in your case, an aborted query). This is different from a test failure. • A failure happens when a test runs successfully but its condition isn't met. [dbt tests outcomes] So your intuition is correct — an

abort query

is a Database Error, not a test failure, meaning the test never actually executed and the underlying data quality is unknown. In Elementary's tables, you can distinguish between these two cases: • Rows with

status = 'error'

in the

elementary_test_results

table represent cases like yours (the test couldn't run), as opposed to

status = 'fail'

which means the test ran and found data quality issues. [removing error results] As for what you can do about it practically, the knowledge sources don't provide a built-in Elementary mechanism to automatically suppress or differently classify

abort query

errors versus genuine test failures. However, a few options are worth considering: 1. Manually remove the erroneous results from

elementary_test_results

by filtering on

status = 'error'

and the relevant date/test name, if you want to clean up the report. [removing error results] 2. Use suppression intervals (

edr monitor --suppression-interval

) to reduce alert noise if these aborts happen repeatedly. [suppression interval] Beyond that, the knowledge sources don't contain enough information to confirm whether Elementary has a native way to differentiate

abort query

errors from actual test failures in its alerting or reporting UI. This might be worth raising as a feature request with the Elementary team. 👍 0 button 👎 0 button ℹ️ Tip: Mention @elementary.ai in the thread for follow-ups or DM the app for 1-to-1 help.