rusackas opened a new pull request, #40030:
URL: https://github.com/apache/superset/pull/40030

   ### SUMMARY
   
   The `Docs Deployment` workflow has been flaky on master for a while. Root 
cause: the deploy step 
(`.github/actions/github-action-push-to-another-repository`) clones 
`apache/superset-site`, drops in the new build, commits, and force-pushes to 
`asf-site` **without rebasing**. When two triggers fire close together — which 
happens routinely because a `docs/**` push fires both the `push` trigger AND 
the `workflow_run` trigger after `Python-Integration` finishes for the same 
commit — the two runs build concurrently and race on the final push. The loser 
dies with:
   
   ```
   ! [rejected]        asf-site -> asf-site (fetch first)
   error: failed to push some refs to 
'https://***@github.com/apache/superset-site.git'
   hint: Updates were rejected because the remote contains work that you do
   hint: not have locally.
   ```
   
   Evidence in the recent run history (same SHA, two runs, one wins, one loses):
   
   | Time | Conclusion | SHA |
   |---|---|---|
   | 2026-05-11 16:50:49Z | success | `96ad2031` |
   | 2026-05-11 16:52:01Z | **failure** | `96ad2031` |
   | 2026-05-11 15:59:09Z | failure | `3e1cfc6d` |
   | 2026-05-11 15:59:16Z | success | `3e1cfc6d` |
   
   Same pattern on `2fe6269c`, `1d1a0e6f`, `ad7075d2`, `6271272e`, `245fffca`, 
etc. — failures interspersed with successes on identical commits.
   
   ### FIX
   
   Add a workflow-level `concurrency` group so only one deploy runs at a time. 
`cancel-in-progress: true` because:
   - The destination repo isn't touched until the very last step (the final 
push), so canceling an in-progress run mid-build is safe.
   - The freshest content is always the right thing to ship — there's no value 
in queueing stale builds behind newer ones.
   
   ```yaml
   concurrency:
     group: docs-deploy-asf-site
     cancel-in-progress: true
   ```
   
   ### BEFORE/AFTER SCREENSHOTS OR ANIMATED GIF
   
   N/A (CI config only).
   
   ### TESTING INSTRUCTIONS
   
   After merge, watch `Docs Deployment` runs on master. Concurrent triggers 
should now see the older run get canceled the moment the newer one starts, and 
the surviving run should push cleanly.
   
   ### ADDITIONAL INFORMATION
   - [ ] Has associated issue:
   - [ ] Required feature flags:
   - [ ] Changes UI
   - [ ] Includes DB Migration (follow approval process in 
[SIP-59](https://github.com/apache/superset/issues/13351))
     - [ ] Migration is atomic, supports rollback & is backwards-compatible
     - [ ] Confirm DB migration upgrade and downgrade tested
     - [ ] Runtime estimates and downtime expectations provided
   - [ ] Introduces new feature or API
   - [ ] Removes existing feature or API


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to