Why is my GitHub Actions workflow stuck in the “queued” state?

[anotheradamibnu]

Select Topic Area

Question

Body

Hi everyone,

I’m trying to run a GitHub Actions workflow on my repository, but it remains in the “queued” state and doesn’t proceed to execution. I’ve already checked the following:
1. There are no pending jobs ahead of it in the queue.
2. I haven’t exceeded the usage limits for my account or repository.
3. The runner for the job is correctly specified and should be available.

Does anyone know what might be causing this issue or how I can debug it further?

Thanks in advance!

[Adamibnualfatahh]

This issue can happen for several reasons. Here are a few things you can check to fix the “queued” state problem:
1. Check Runner Availability
Ensure that the runner you specified in your workflow file is available and not occupied by other jobs. You can check the status of your self-hosted runners in the repository settings under Actions > Runners. If you’re using GitHub-hosted runners, they might be busy with other jobs, especially during high demand periods.
2. Job Dependencies
If your workflow includes jobs that depend on others, make sure those jobs have completed successfully before the queued job starts. Sometimes, if a job dependency fails or hasn’t started, it can cause subsequent jobs to stay in the “queued” state.
3. Check Usage Limits
If you’re on a free plan or have hit usage limits, this can cause jobs to be queued until capacity is available. You can review the usage limits for GitHub Actions in your account settings to make sure you’re within the limits.
4. Cancel Stuck Jobs
Sometimes a stuck job can cause other jobs to get queued indefinitely. If this is the case, try canceling the queued job and re-running it. This can sometimes resolve the issue.
5. GitHub Status
Lastly, check GitHub’s status page to see if there are any ongoing issues with GitHub Actions or infrastructure that might be causing delays.

Let me know if any of these steps help!

[mecodeatlas]

Thanks for posting in the GitHub Community, @anotheradamibnu!

We're happy you're here. You are more likely to get a useful response if you are posting your question in the applicable category, the Discussions category is solely related to conversations around the GitHub product Discussions. This question should be in the Actions category. I've gone ahead and moved it for you. Good luck!

[LiamSwayne]

I am having a similar issue but all my actions are frozen as queued across all repos. Any reason this may be happening? It seems to have happened overnight and is not the result of any code changes.

[user-Roma]

@LiamSwayne , having the same issue today evening, thats why im here.
Looks like smthng on github side

[user-Roma]

I tried to rerun pipelines for working in the past deployments and they also having status Queued.
I thought that maybe i could reach some limits but it was first run during last 3 days..

[petedanielsmith]

mine was queued for an hour, i quit it thinking it was my fault and repushed and its still an hour and a half queued now. def an issue their end, although status page still shows no issues anywhere

[LiamSwayne]

mine started working 25 seconds ago

[user-Roma]

for me still not working

[stampd-dev]

not working again for me today

[LiamSwayne]

@github please look into this it has crashed production for me and probably a lot of other people that haven't noticed yet

[NikitaVelixJob]

bump, same issue Queued

[petedanielsmith]

I'm well below usage limits, I tried restarting, it's clearly an issue as it's happening to multiple people. The status page is no help as just says everything is fine. Now been over 12 hours stuck in queue and GitHub aren't even reporting an issue

[NikitaVelixJob]

i restarted my job and now its worked

[petedanielsmith]

I did that when it first got stuck for an hour. Just reluctant to try again if it means I get stuck in another 12 hour queue at the back again 😀 maybe I'll give it a go

[petedanielsmith]

Finally it's working again!

[user-Roma]

@iampraveen08 , my pipelines also started working working by itself few minutes ago, after another page refresh

[Prath-0506]

1. Check GitHub Status:Visit status.github.com to see if GitHub Actions runners are experiencing outages or high load.
2. Are you using GitHub-hosted runners (like ubuntu-latest)?
   If yes — they should auto-scale.
   Are you using a self-hosted runner?
   Make sure it’s online and shows as Idle in your repo’s Settings → Actions → Runners.
3. Double-check your runs-on in the workflow YAML. It must match the runner’s labels exactly (for example: runs-on: [self-hosted, linux]).
4. If it’s a PR from a fork, do you see an “Approve and run” button in the Actions tab?.Someone with write access must approve it.
5. Are there other jobs ahead in the queue?. Sometimes concurrency or limited runners can create a hidden backlog.
6. Visit your repo/org’s Actions Usage settings to see if you hit free minutes or storage limits.
7. Do you use concurrency in your YAML?.If so, a previous run in the same group might be blocking new runs.
8. Is your runs-on or matrix syntax valid?. Even small typos can prevent a job from matching any runner.
9. If it’s an org repo, see if there are policies restricting runner groups. Example: your repo may not have permission to use certain runners.
10. Click the workflow run → click the queued job → see if you get hints like “Waiting for runner to come online.”
11. Sometimes the runner process has crashed or disconnected.Stop & restart it, or re-register if needed.
12. Create a new workflow that just echoes Hello World. This helps rule out YAML issues.
13. Temporarily change runs-on to a different OS/version. Example: ubuntu-latest instead of ubuntu-22.04.
14. For private repos, confirm you have Actions minutes left.Free accounts have limited free minutes.
15. If you’ve tried everything and it stays queued for hours, open a support ticket.
16. Most commonly, it’s either:
    • The runner is offline/unavailable,
    • The labels don’t match,
    • Or you need to approve a workflow from a fork.

[ghost]

Similar issue. pages-build-deployment is stuck for more than 6 hours
https://github.com/lingonas/lingonas.github.io/actions/runs/16104288239

Tried "Cancel run", but it returns "Failed to cancel workflow."

[F3rrY]

Same here, right now

[abrehamgezahegn]

Experiencing the same issue here.

[w4ss1mIBM]

Experiencing the same issue here.

[nghhoese]

same issue here suddenly some workflows wont run anymore and stay queud for forever canceling and restarting dont seem to work.
Thought i was going crazy before finding this thread.

[abinasharma001]

💬 Answer:

Hi! You're doing the right checks already — thanks for sharing the context. If your GitHub Actions workflow is stuck in the “queued” state, but nothing obvious is blocking it, here are some additional things to verify:

---

✅ Troubleshooting Checklist

1. 🧠 Check for Runner Availability

Even if you’ve specified a runner, make sure that:

• The runner is online and idle (for self-hosted).
• You’re targeting the correct label (ubuntu-latest, self-hosted, etc.).
• For enterprise or org-hosted runners, check if the runner group has permission for the repo.

Tip: Go to Repository Settings → Actions → Runners to check runner status.

---

2. 📦 Workflow Concurrency Limits

Check for concurrency settings in your workflow YAML. If you’ve configured it like:

concurrency:
  group: my-job
  cancel-in-progress: true

...then another run might be blocking it without being obvious.

---

3. 🔒 Repository Permissions or Secrets

If the workflow uses secrets or environments with protection rules, GitHub might delay or pause the job until approval or access is granted.

---

4. ⚙️ GitHub Actions System Issues

Check:

• [GitHub Status](https://www.githubstatus.com/)
• [Actions incident history](https://www.githubstatus.com/history)
  GitHub sometimes has runner-related incidents that can delay job starts.

---

5. 🪵 Enable Debug Logging

Add the following to your workflow or repo secrets for more logging:

ACTIONS_RUNNER_DEBUG: true
ACTIONS_STEP_DEBUG: true

---

🛠️ Still Not Running?

If it stays queued for >5–10 minutes with no logs or movement:

• Cancel and re-run the job.
• Try triggering it from another branch.
• Post a screenshot or YAML snippet here for more targeted help.

---

Hope this helps you get unstuck! Let me know what your runs-on and concurrency values look like if you're still blocked.

[LiamSwayne]

Bro this is AI-generated lmao pls stop. For people wondering why this is happening it's probably actions being over capacity. Even on the pro plan I was not able to get anything to run.

[petedanielsmith]

Think it was definitely a bug. I found if I replied on a push triggering it that it went to queue and stayed there, sometimes for like 12 hours, but if I cancelled that job and reran failed jobs it worked but a fresh push would go back in the stuck in queue status. Rerunning just 1 failed job seemed to also get it to pick up other jobs I had stuck in queue too. Not tried again today though

[LiamSwayne]

Yea that sounds like a bug

[grant-zukowski-xumo]

For me I was gone for awhile and I had my runner version os to ~ubuntu 20. There was no error but it hung on queue. I bumped the runner version to 22.04 and that got it unstuck.

I hope this helps someone.

[GajaNen]

Yep, I was in the same situation and changing to 22.04 worked., thanks!

[akochepasov]

This is the solution. They stopped supporting old OS runners

[ajimix]

This also worked for me 🙏

[pipedexplumbing]

Same issue.

[james-harlyy]

Happening for me, starting today. Not a usage issue, we've used less for 5% of our Enterprise plan usage. All the workflows are stuck in "queued" (for about an hour now)

[s-titov]

Same here, can't deploy to prod, just infinite Queued. I've tried re-run but it didn't help.

[Sewer56]

I ran into this problem recently.
When I contacted GitHub Support, the person at the other end was helpful;
and they also let me know of a certain API after helping me out:

• Force cancel a workflow run (See the docs for more details)

There's a chance this might help you folks.

[sramekmichal]

If you get something like:

{
  "message": "Cannot cancel a workflow re-run that has not yet queued.",
  "documentation_url": "https://docs.github.com/rest/actions/workflow-runs#force-cancel-a-workflow-run",
  "status": "409"
}

Just run in your repo:

git commit --allow-empty -m "trigger re-run"
git push

It is gonna skip the previous jobs in stuck state and run fresh job.

[L0g0ff]

This works! Thanks!

[jalbarranelectronic]

Some years later, this tip saved my life. Thanks!

[tolzy88]

I owe you a beer!

[Vishnupriya-Anupindi]

I am having the same problem now, don't know what to do. I tried cancelling the workflow and to Re-run all jobs. I added the following to the yml file and changed the node version from 16.x to 20.x but even that job is Queued.
concurrency:
group: build
cancel-in-progress: true

[jellytha]

[jellytha]

[Bluesboy]

Same issue here, can't even cancel the workflow, UI gives says "Failed to cancel workflow.", CLI says "Cannot cancel a workflow run that is completed". None of methods in the accepted answer work.

[Gutza]

I am having the same problem. I have four instances of the same action on a GitHub pages repo, all stuck in Queued state, with Failed to cancel workflow. if I try to cancel them.

For my use case, it's obvious this is a bug: this is a simple Jekyll build-and-deploy action with three tasks:

1. Build the Jekyll website
2. Report the build status
3. Deploy the website

The dependencies are (1) -> (2) and (1) -> (3) – notably, not (1) -> (2) -> (3).

All four instances have tasks (1) and (2) completed successfully, and task (3) completed, timed out, or canceled – and yet the actions are all Queued!

So this is the perfect Catch-22:

• Attempting to cancel the run doesn't make sense, since no task is running (nor is it really queued to run, since all actions are in final states) – so the error "Failed to cancel workflow." is legitimate (there's nothing to cancel);
• Starting a new action fails, because there are "queued" actions which also attempt to deploy to the same website;
• However, because of the bug, timed out deployment tasks don't error the action instance, they leave it queued indefinitely.

[Gutza]

This fixed it for me: https://github.com/orgs/community/discussions/49074#discussioncomment-15489079