Timedelta dag backfill dates don't align with normal schedule
- Dominant language
- Python
- Stars
- 46.9k
- Forks
- 17.8k
- Avg merge
- 2d 10h
- Merged PRs (30d)
- 483
Description
### Body
With a timedelta-scheduled dag, the logical dates that you get are relative to the start-date you provide when running the backfill command.
**This has been the behavior since 2.x**
So, e.g. suppose your dag has schedule of timedelta(days=1). Then, running this
```
airflow dags backfill test_backfill_dag__timedelta --start-date '2025-03-10 03:25:26' --end-date 2025-03-13 --dry-run
```
Will produce these runs:
```
"2025-03-10 03:25:26+00:00"
"2025-03-11 03:25:26+00:00"
"2025-03-12 03:25:26+00:00"
```
But if the schedule was cron schedule, then it would be what you would expect -- just the "partitions" that are within the range requested.
### Committer
- [x] I acknowledge that I am a maintainer/committer of the Apache Airflow project.
Contributor guide
Research direction
Start with the `airflow dags backfill` command and the timedelta-scheduled DAG behavior described in the examples. Compare its dry-run logical dates with the requested range and with cron scheduling; done means timedelta backfills produce dates aligned with the normal schedule.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- data-engineering
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100