Hi,

With no further discussion, consensus reached

Thanks!

Andrew and Teghveer

On Tue, 21 Jul 2026 at 12:48, Jung-Hyun Andrew Kim <
[email protected]> wrote:

> Hi all,
>
> This is a follow up to the discussion thread:
>
> https://lists.apache.org/thread/l1b1wkxs69omjsb8cwdryr2oydjmn3yg
>
> To reiterate quickly, the /health endpoint does not reveal meaningful
> information for distributed environments because it reports the system as
> healthy if at least one of the following components is running and
> heartbeating: the scheduler, DAG processor, or triggerer. This fails to
> show meaningful status for distributed environments.
>
> Because discussion has mostly settled and the consensus is that Teghveer
> and myselfs proposal is good to move forward I would like to call a lazy
> consensus on the following design:
>
> Design Doc:
> https://docs.google.com/document/d/1yWO-0YL6TNpABLwXoKy3YGVFZGnr4CE0Q1iZ_rpNmBo/edit?usp=sharing
>
> Example PR: https://github.com/JH-A-Kim/airflow-fork/pull/4
>
> The Lazy consensus will be reached in 72 hours on July 24th 13:00 PST.
>
> Big thanks to Dennis, Vincent, and Niko for their feedback!
>
> Thanks,
>
> Andrew
>
>

Reply via email to