GitHub Actions Runner No Space Left On Device
Get a free runner storage preview first: largest buckets, a safe starter cleanup YAML, and the confirm-first paths that should not be deleted blindly. If the failure still needs judgment, send one cleanup request and continue to the SafeDisk AI deletion trial only if the free check is not enough.
Find what the runner can delete.
Leave your email now; the scan summary or failing job link can follow after the first reply. We send the free SafeDisk AI deletion trial step only if deletion risk is still unclear.
Run the read-only Action
Run the free CI storage Action first. If the job still needs review, send one request for the free SafeDisk AI deletion trial next step.
- name: SafeDisk CI storage scan
if: always()
uses: liuminsheng3/safedisk-ci-scan-action@v1
What the scan checks
- Runner disk overview, workspace, temp, tool cache, current job paths, and runner image metadata.
- Linux runner pressure points: `/home/runner/actions-runner`, `_work`, `_diag`, `_tool`, `/tmp`, `/var/lib/docker`, and Docker build cache.
- macOS runner pressure points: Xcode, DerivedData, Archives, CoreSimulator, SwiftPM, CocoaPods, Homebrew, npm, pnpm, and Gradle caches.
- Large workspace files and common build outputs that grow silently across repeated CI jobs.
How to interpret no-space failures
A CI runner can be out of usable space even when the root cause is not obvious from the failing command. `no space left on device` can be triggered by Docker overlay growth, a build cache, a test artifact directory, `/tmp`, Xcode simulator runtimes, or a self-hosted runner that never rotates old workspaces.
Usually safe to automate
- Per-job temp files after the job finishes.
- Known rebuildable package caches with a size cap.
- Old workflow artifacts copied into runner-local scratch paths.
Review before deleting
- Docker images, build cache, and volumes used by later jobs.
- Shared tool caches on self-hosted runners.
- Xcode runtimes, SDKs, and simulator state required by pinned builds.
Do not touch blindly
- Runner registration, service files, credentials, and `_diag` logs needed for incident review.
- Persistent Docker volumes that may contain databases or test fixtures.
- Shared build outputs that are reused outside the current job.
Local CI preview
Preview the runner policy before sending anything.
Upload or paste the generated CI scan. The preview runs in your browser and stores the scan only for request prefill.
Paste a short storage summary to preview the runner cleanup plan.
Close the incident with a policy
Copy the read-only check first. Request the free SafeDisk AI deletion trial only if the output, app state, or cleanup boundary is still not obvious.
Need a delete / confirm / protect answer for this runner?
Submit the form first; the failing job link can follow. We check whether free guidance is enough before asking for the free SafeDisk AI deletion trial.