You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@airflow.apache.org by "Kaxil Naik (Jira)" <ji...@apache.org> on 2019/12/16 15:52:00 UTC
[jira] [Updated] (AIRFLOW-4081) Task failures with "Task instance's
dagrun did not exist: Unknown reason."
[ https://issues.apache.org/jira/browse/AIRFLOW-4081?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Kaxil Naik updated AIRFLOW-4081:
--------------------------------
Labels: easyfix (was: documentation easyfix)
> Task failures with "Task instance's dagrun did not exist: Unknown reason."
> ---------------------------------------------------------------------------
>
> Key: AIRFLOW-4081
> URL: https://issues.apache.org/jira/browse/AIRFLOW-4081
> Project: Apache Airflow
> Issue Type: Bug
> Components: scheduler
> Affects Versions: 1.10.2
> Environment: I'm using the latest from PyPI, but this bug was tricky enough to identify that seems likely to be in master too.
> Reporter: Tylar Murray
> Priority: Minor
> Labels: easyfix
> Original Estimate: 168h
> Remaining Estimate: 168h
>
> Tasks will fail with `Task instance's dagrun did not exist: Unknown reason` and show no log, returning only `{{*** Failed to fetch log file from worker. Invalid URL 'http://:8793/log/dag_id/task_id/2019-03-13T11:21:00+00:00/1.log': No host supplied` when:}}
> # {{dag `max_active_runs=1` and `catchup=False`}}
> # task remains queued (or is running?) when a 2nd dag is triggered
> This error state had me confused about "random" failing tasks, but the pattern now seems clear and the error should be reproducible in a clean environment.
> A fix should include a more informative error message and perhaps the task state should be set to "skipped" rather than "failed".
--
This message was sent by Atlassian Jira
(v8.3.4#803005)