Our dbt tests fail every night and send us an email notifying us of the failure. This is not a relevant notification; we know there are test failures and that they will continue for the foreseeable future. We need a way of extracting the relevant signal from these failing tests, so that our notifications can be informative.
One idea is to extract and store the number and type of failures every day, and then populate some kind of dashboard including things like a burndown chart showing increase or decrease in number of test failures over time. There's some flexibility and uncertainty here, so we would probably do well to start with a design or a proof of concept.
Our dbt tests fail every night and send us an email notifying us of the failure. This is not a relevant notification; we know there are test failures and that they will continue for the foreseeable future. We need a way of extracting the relevant signal from these failing tests, so that our notifications can be informative.
One idea is to extract and store the number and type of failures every day, and then populate some kind of dashboard including things like a burndown chart showing increase or decrease in number of test failures over time. There's some flexibility and uncertainty here, so we would probably do well to start with a design or a proof of concept.