Repository navigation
Stop reporting components as degraded after an instance restart - #73222
Merged
vincbeck merged 2 commits intoSep 18, 2026
Merged
Conversation
vincbeck
requested review from
ephraimbuddy,
jason810496,
jedcunningham,
pierrejeambrun and
rawwar
as code owners
September 15, 2026 19:20
boring-cyborg
Bot
added
area:API
vincbeck
force-pushed
the
vincbeck/health_detailed_status_coverage
branch
from
September 16, 2026 14:35
712ddb0 to
5c8f1a1
Compare
vincbeck
requested review from
bbovenzi,
bugraoz93,
choo121600,
dheerajturaga,
guan404ming,
henry3260,
potiuk,
ryanahamilton,
shubhamraj-git and
vatsrahul1001
as code owners
September 16, 2026 14:35
Contributor
Author
|
@pierrejeambrun I would love to get your opinion on this one :) |
pierrejeambrun
left a comment
pierrejeambrun
left a comment
Member
There was a problem hiding this comment.
Nice one.
Overall looking good, just a few nits / questions.
vincbeck
force-pushed
the
vincbeck/health_detailed_status_coverage
branch
from
September 17, 2026 20:18
7996e8c to
623305d
Compare
The health endpoint's ``detailed_status`` compared live job rows against all unfinished job rows, but a job row is only marked finished by a cooperative shutdown. An instance lost to SIGKILL, an out-of-memory kill, or a node eviction leaves its row unfinished forever, so restarting any scheduler, triggerer, or Dag processor made the component look permanently ``degraded`` even though the deployment was fully healthy. Those same abandoned rows were also listed in ``instances``, advertising hosts that no longer exist. The question ``detailed_status`` should answer is whether every part of a component's work has a live instance covering it, which needs the expected set of work to be declared rather than inferred from whatever rows happen to be in the table. Dag bundle configuration already declares it for the components that partition their work, and schedulers do not partition theirs at all.
vincbeck
force-pushed
the
vincbeck/health_detailed_status_coverage
branch
from
September 18, 2026 13:52
623305d to
e22cfc0
Compare
1 task
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The health endpoint's
detailed_statuscompared live job rows against all unfinished job rows, but a job row is only marked finished by a cooperative shutdown. An instance lost to SIGKILL, an out-of-memory kill, or a node eviction leaves its row unfinished forever, so restarting any scheduler, triggerer, or Dag processor made the component look permanentlydegradedeven though the deployment was fully healthy.The question
detailed_statusshould answer is whether every part of a component's work has a live instance covering it, which needs the expected set of work to be declared rather than inferred from whatever rows happen to be in the table. Dag bundle configuration already declares it for the components that partition their work, and schedulers do not partition theirs at all.Was generative AI tooling used to co-author this PR?
{pr_number}.significant.rst, in airflow-core/newsfragments. You can add this file in a follow-up commit after the PR is created so you know the PR number.