Hi team,
I am running into a Space runtime issue that looks platform-side rather than application-side.
My Space appears to be stuck in a broken runtime-control state.
Symptoms:
- Build logs complete successfully.
- Factory reboot from the Space UI returns 503.
- Restarting via API also fails with:
{“error”:“Something went wrong when restarting this Space.”} - Changing Secrets or Variables does not trigger a new runtime.
- The Space UI does not show “Duplicate this Space”.
Because of this, I cannot roll out updated code even though the image itself builds successfully.
What I tested:
- Normal UI restart
- Factory reboot from the UI
- Changing Secrets / Variables to trigger a runtime restart
- Restart via API:
POST /api/spaces/<repo_id>/restart - Direct network access to huggingface.co from my local machine is fine
Observed behavior:
- Build phase succeeds
- Runtime restart/control actions fail
- The Space remains stuck on the old runtime state
Example API result:
{“error”:“Something went wrong when restarting this Space.”}
Request ID from the UI error page:
Root=1-69b97e50-53d7be5809cc76cd7ec31f5e
Could you please check whether this Space is stuck on the platform side and help reset or recover its runtime state?
Thanks.
Additional details:
- This affects deployment of new code, not just the currently running app.
- The issue is reproducible both from the UI and from the API.
- Local HTTPS access to https://huggingface.co works normally, so this does not look like a local connectivity problem.