Pages publish latency regressed from ~10s to >600s (Actions deploy), causing deploy-pages timeouts #204128
Replies: 11 comments
|
💬 Your Product Feedback Has Been Submitted 🎉 Thank you for taking the time to share your insights with us! Your feedback is invaluable as we build a better GitHub experience for all our users. Here's what you can expect moving forward ⏩
Where to look to see what's shipping 👀
What you can do in the meantime 💻
As a member of the GitHub community, your participation is essential. While we can't promise that every suggestion will be implemented, we want to emphasize that your feedback is instrumental in guiding our decisions and priorities. Thank you once again for your contribution to making GitHub even better! We're grateful for your ongoing support and collaboration in shaping the future of our platform. ⭐ |
|
Same here. |
|
Same problem, if there's any update let me know |
|
Same problem |
|
Same problem |
|
Same |
|
Yup, same problem here. |
|
Same here, on a second unrelated repo — Queue duration below is measured from
A few variables our two reports jointly rule out:
The retry conflict has a knock-on effect worth flaggingEach timed-out attempt holds the So retrying actively makes it worse. Probably worth waiting it out instead. No user-side mitigation is availableRaising the const MAX_TIMEOUT = 600000
this.timeout = !timeoutInput || timeoutInput <= 0 ? MAX_TIMEOUT : Math.min(timeoutInput, MAX_TIMEOUT)With publish latency now sitting right at ~600s, whether a deploy passes is a coin flip on the 10-minute boundary — my 11:05 run cleared it by 8 seconds, and the two after it did not. githubstatus.com is still all-green, with no Pages incident recorded. |
|
Confirming the same regression on a repo using the branch-based publishing source All times UTC, 2026-08-06. Last healthy deployment: 2026-08-05 19:05:16Z (run took 47s).
That last one is exactly 600s in Worth flagging a cascade that makes this worse than a slow deploy, in case it
githubstatus.com still reports Pages and Actions as operational, which is Net effect: the site keeps serving the last successful build (no outage), but |
|
Same here |
Uh oh!
There was an error while loading. Please reload this page.
🏷️ Discussion Type
Bug
💬 Feature/Topic Area
Pages
Body
Since some time between 2026-08-05 13:07 UTC and 2026-08-06 08:45 UTC, GitHub Pages publishes for my repository have gone from about 10 seconds to over 10 minutes. The repo is
dmellok/tesserae, public, deployed from Actions (build_type: workflow,actions/deploy-pages@v5) to the custom domaindocs.tesserae.ink.This has made the site unpublishable, because
actions/deploy-pagesclamps its status-polltimeoutto a maximum of 600000 ms. Any publish that overruns 10 minutes gets aborted by the action, which then cancels the deployment.Duration of the
deployjob across recent runs, all onmain, all deploying a near-identical artifact:A representative failure, from run 31104001281:
The automatic retry fails immediately because the previous deployment for the same commit is still settling. In one earlier run it returned HTTP 400: "Deployment request failed ... due to in progress deployment. Please cancel
<sha>first or wait for it to complete."What I have ruled out:
buildjob is unchanged at roughly 30 to 40 seconds throughout. The whole regression is in the publish step.github-pagesenvironment's branch policy permitsmain, and the HTTPS certificate fordocs.tesserae.inkisapproved, expiring 2026-09-25.Two questions. Has anyone else seen Pages publish latency jump by this order of magnitude in the last day or so, on Actions-based deployments? And when a publish is aborted mid-flight by the action's timeout, is anything left server-side that needs clearing before the next deployment can proceed, or does it settle on its own?
All reactions