thesuperzapper commented on issue #34405: URL: https://github.com/apache/airflow/issues/34405#issuecomment-2356954268
@potiuk @kaxil @eladkal @ferruzzi I think this is a show-stopping issue for Open Telemetry integration in Airflow. @potiuk said that he thinks this expected behavior (see: https://github.com/apache/airflow/pull/40286#issuecomment-2205854126), but I strongly disagree for the following reasons: 1. This is not the behavior of StatsD integration. That is, StatsD being down does not cause all tasks across the cluster to fail. 2. The telemetry being successfully sent does not change the fact that my task may have succeeded in making some external change. For example, if my task was loading data into a table, I __really don't__ want to do it twice because OpenTelemetry was down and so marked the task a "failed". At the very least, we need to make this a config, but I honestly think the default value should be "warn and continue", rather than "fail the task" as it's so dangerous in the current state. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
