j-adamczyk opened a new issue #18698: URL: https://github.com/apache/airflow/issues/18698
### Apache Airflow version 2.1.4 (latest released) ### Operating System Alpine Linux 3.13 ### Versions of Apache Airflow Providers apache-airflow-providers-ftp==1.1.0 apache-airflow-providers-imap==1.0.1 apache-airflow-providers-sqlite==1.0.2 ### Deployment Docker-Compose ### Deployment details Celery on Kubernetes, with logging on AWS S3. ### What happened Tasks fail randomly (absolutely randomly, any task can fail with this exact message at any moment) with log: ``` *** Falling back to local log *** Log file does not exist: /opt/airflow/logs/reporting_eventbrite_attendees/get_eventbrite_events/2021-10-03T00:00:00+00:00/1.log *** Fetching from: http://:8793/log/reporting_eventbrite_attendees/get_eventbrite_events/2021-10-03T00:00:00+00:00/1.log *** Failed to fetch log file from worker. The request to ':///' is missing either an 'http://' or 'https://' protocol. ``` This is caused by UI server fetching logs for the worker, not by worker itself (it does not show anything in the logs). So the whole DAG may fail due to Airflow UI problem, not due to any programmatic error. ### What you expected to happen If the log is not found, just show "No logs could be fetched" or similar message. Mark task as success if no error was thrown in it. Continue the DAG execution. ### How to reproduce This is not reproducible even in our case. This fails randomly, but for about ~1/3rd of the tasks. However, for all log fetching-related problems any UI problem (not only this particular one) should not result in task failure. ### Anything else Problem appears randomly, with no apparent changes. There are no particular spikes in usage of CPU, RAM or anything else. Other DAGs that run concurrently with the failing task work properly. ### Are you willing to submit PR? - [X] Yes I am willing to submit a PR! ### Code of Conduct - [X] I agree to follow this project's [Code of Conduct](https://github.com/apache/airflow/blob/main/CODE_OF_CONDUCT.md) -- 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]
