edit: figured it out...

detection_delay

does not ignore the 0 row counts. It stops the test at x periods back from the execution time. Adding a

detection_delay

after already having executed

dbt_test

will not work to clear the failure because the test already ran and stored results. ======= ORIGINAL MSG ======== Trying to understand how the new detection_delay works. I have a table that often shows 0 volume on 1 - 2 days in the past. So I thought I'd set up detection_delay so I don't see failures if the volume fails in the last 2 days. But the test still fails