GitHub Actions Runs Stuck Queued: A Challenge for Software Developer Overview

The Persistent Puzzle of Stuck GitHub Actions Runs

In the fast-paced world of continuous integration and delivery, reliable automation is paramount. However, even robust platforms like GitHub Actions can encounter perplexing issues. A recent discussion on the GitHub Community forum highlights a particularly stubborn bug where workflow runs become 'stuck queued,' defying standard cancellation and deletion attempts. This scenario presents a critical challenge for maintaining a clear software developer overview of project status and progress.

The Core Problem: Uncancelable, Undeletable Queued Runs

The user, SakenW, reported two specific GitHub Actions runs for their transhub-ecosystems-source-executor repository that were stuck in a 'queued' state. Despite being created days earlier (October 20, 2026), their status remained 'queued' with a conclusion: null and unchanged timestamps. Key characteristics of these problematic runs included:

• Zero jobs across all attempts.
• Zero check runs and zero artifacts.
• The matching self-hosted runner was online and idle.
• The workflow used a fixed concurrency group with cancel-in-progress: false.

Failed Self-Service Attempts

SakenW meticulously documented several attempts to resolve the issue, all of which met with unexpected errors, demonstrating the depth of the problem:

• gh run cancel: For both runs, the CLI command returned "Cannot cancel a workflow run that is completed," directly contradicting the 'queued' status.
• API POST /cancel and /force-cancel: Direct API calls resulted in HTTP 409 Conflict errors, stating, "Cannot cancel a workflow run that has not been queued yet." This presented another conflicting status.
• API DELETE: An attempt to delete one of the runs (39451480378) with repository-owner authorization resulted in an HTTP 403 Forbidden error, "Could not delete the workflow run." The request ID was 9AED:341381:E06F73:142EBDD:6ADCEF97.

Impact and Lack of Resolution

The persistence of these phantom queued runs has practical implications. The user's dispatcher system conservatively treats queued runs as active, which prevents further automatic processing, effectively blocking the workflow. The discussion highlights the difficulty in obtaining direct technical support for GitHub Actions bugs, with the Support portal stating "Technical support not included" for this category.

The only response to the discussion was an automated "Product Feedback Has Been Submitted" message, which, while acknowledging the feedback, offered no immediate solution or workaround. This leaves developers like SakenW in a challenging position, awaiting intervention from GitHub staff to reconcile or clear these stale states.

Lessons for Developer Productivity

This incident underscores the importance of robust error handling and recovery mechanisms within CI/CD platforms. For any software developer overview, understanding potential points of failure and the available recovery paths is crucial. While GitHub Actions generally offers excellent reliability, edge cases like these can significantly impact development velocity and require direct platform intervention. It also highlights the need for clear support channels for critical infrastructure issues.

The community continues to be a vital resource for reporting and discussing such challenges, hoping for official resolutions that enhance the overall developer experience.