Workflows on my fork are repeatedly disabled after being re-enabled #213366
Replies: 2 comments
|
Hi @yunnysunny — yes, this is expected behaviour in two documented cases, but the repeat disabling on a short cycle needs a diagnosis, not just another re-enable. I run a scheduled workflow every 6 hours on a repo I own outright, and I've answered a few schedule questions here, so I've spent time in these docs — but one honest caveat first: my own schedules don't run from a fork, so the fork part below is docs plus what your public repo shows, not my own fork experience. 1. What disables them — the two documented causes. From GitHub's docs ("Disabling and enabling a workflow"):
So the fork rule explains the first disable: when the fork was created, its schedules started disabled. The 60-day rule explains a later one: a public repo with no activity for 60 days loses its schedules. Re-enabling is What doesn't fit neatly: I checked your repo just now. All three workflows currently report 2. How to tell when and why — check the state before you re-enable. The workflow API exposes the reason as the The values are On audit logs: your repo is owned by an organization, so org owners can open the organization audit log — but what I could verify there is 3. The reliable fix — move the schedule out of the fork. If the generated data matters to you, the durable answer is to run the schedule in a repo you own outright, not in a fork: the fork default-disabled rule simply doesn't apply there, and only the 60-day rule remains. Two ways to do that without losing the setup: create a standalone repo with the same workflow files (the upstream repo is GPL-3.0, so keep the licence), or keep the fork as the data mirror and put the schedule in your own repo, triggering the update there. If it has to stay in the fork, the only thing that keeps the 60-day rule away is activity — any push resets the clock, and a scheduled upstream sync that lands commits counts. But that only covers the 60-day cause; if the next disable shows If you post the |
|
Update: the issue is still present, and I have not re-enabled the workflows. On November 18, 2026 at approximately 01:54 UTC, the repository's Actions page still showed “Workflows on this fork have been disabled” and offered the button to re-enable them. At the same time, the workflows API reported all three individual workflows as
The most recent run was a successful scheduled run created on November 6 at 21:28 UTC: https://github.com/whyun-pages/iptv-sources/actions/runs/42297008158 This appears to be a discrepancy between the fork-level Actions page and individual workflow API states. The exact disablement time and cause are unknown. Is there another API or audit record that exposes the fork-level state or explains why this notice returned? |
Uh oh!
There was an error while loading. Please reload this page.
🏷️ Discussion Type
Question
💬 Feature/Topic Area
Schedule & Cron Jobs
Discussion Details
Hi GitHub Community,
I maintain a forked repository: https://github.com/whyun-pages/iptv-sources
The repository uses scheduled GitHub Actions workflows to keep its generated data up to date. I have encountered the following notice more than once:
When this happens, the scheduled updates stop. I previously re-enabled Actions, but the same notice appeared again later.
I understand that GitHub disables workflows on forks to prevent unnecessary compute usage. However, I would like to understand why they are being disabled repeatedly and how to prevent it.
I can provide additional details or timestamps if needed. Thanks for your help!
All reactions