GitHub user starlight1805 created a discussion: Intermittent "recorded hostname
does not match instance hostname" errors with multiple active schedulers
(LocalExecutor, Airflow 2.7.0)
We are running into an intermittent issue with Airflow 2.7.0 where many tasks
fail at once with the following error:
**[2025-09-23, 06:23:08 IST] {local_task_job_runner.py:253} ERROR -
The recorded hostname hostname1.com does not match this instance's hostname
hostname2.com**
Architecture
- Airflow 2.7.0
- LocalExecutor
- MySQL metadata DB
- All logs, DAGs, and configs stored on a NAS mount (shared between nodes)
- Active-active schedulers enabled (3 schedulers total)
Cluster setup:
- Node 1 → webserver + scheduler
- Node 2 → scheduler only
- Node 3 → scheduler only
Observed behavior
- Normally jobs run fine across all three nodes.
- About once every two weeks, ~50 DAGs fail simultaneously with the above
hostname mismatch error.
- The mismatch is always the same pattern:
- This happens even if the job is actually running on Node 1, Node 2, or Node 3.
- After this window, jobs continue to run normally until the next occurrence.
GitHub link: https://github.com/apache/airflow/discussions/56024
----
This is an automatically sent email for [email protected].
To unsubscribe, please send an email to: [email protected]
