db clean --skip-archive still copies deletion candidates into temporary archive tables prior to deletion. #75966

Under which category would you file this issue?

Airflow Core

Apache Airflow version

3.3.2

What happened and how to reproduce it?

On any Airflow 3.3.2 metadata database containing log rows older than the cutoff, run:

  airflow db clean \
    --tables log \
    --clean-before-timestamp '2026-10-28T00:00:00+00:00' \
    --batch-size 1 \
    --skip-archive \
    --yes \
    --verbose

You can see the following per-batch output:

1. Moving data to table _airflow_deleted__log__... messages, and
2. CREATE TABLE ... AS SELECT ... debug SQL output.

Despite --skip-archive being present, each batch copies rows into a staging table before deleting them from log, then drops the staging table. Because of this, checking for archive tables after cleanup will not reveal the intermediate copies, and it also slows down cleanup when skip_archives is enabled.

What you think should happen instead?

With --skip-archive, deleting eligible rows should not create or populate any archive/staging tables.

Operating System

ubuntu 24.04, also MacOS 26.6.2

Deployment

No response

Apache Airflow Provider(s)

No response

Versions of Apache Airflow Providers

No response

Official Helm Chart version

No response

Kubernetes Version

No response

Helm Chart configuration

No response

Docker Image customizations

No response

Anything else?

Only realized after filing this issue that there is currently an open candidate solution for it: #74113

Are you willing to submit PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct