a month ago
Hi. Since about 15:00 ET on 6 Oct 2026 (19:00 UTC), every deploy of my web service has failed at the deploy step with: "This deployment expects a volume mounted at /data, but the service has no volume attached. Railway did not create the container."
The volume is attached: railway volume list shows it on the service, mounted at /data, status Ready. Nothing changed in the service settings between the last good deploy and the failures; the deploy config is identical. Your built-in Diagnose on the latest failed deploy calls it an Infrastructure Error: the deploy-time check reported no volume at /data even though the service config lists one.
Builds succeed every time; the deploy stops at the volume check. The dashboard also says "There is no active deployment for this service" even though the last good deployment is still serving traffic.
The service is linked to this thread, and I can share project and deployment IDs privately. The volume holds data I need to keep, so I'd rather not detach and reattach it unless you confirm that's safe. Could you check whether the volume-to-service link is out of sync on your side?
Thanks,
Spencer
3 Replies
a month ago
I’m seeing the exact same issue on a separate Railway project, also in sfo.
My volume is definitely attached and healthy:
- Mount: /app/data
- Status: Ready
railway status, railway volume list, and the failed deployment manifest all confirm the /app/data volume is attached.
I also managed to isolate the issue:
- With my normal Pre-deploy Command (python -m alembic upgrade head) enabled, deployment fails immediately with the same “expects a volume mounted at /app/data, but the service has no volume attached” error.
- I removed the Pre-deploy Command and temporarily moved the migration into the Start Command. The exact same service/image/volume deployed successfully.
- I then restored the original Pre-deploy Command and removed the custom Start Command, and the same volume error immediately returned.
Other services in the same Railway project also have volumes mounted at /app/data and are deploying successfully; they do not have a Pre-deploy Command.
So this appears to be an infrastructure regression involving volume-backed services with Pre-deploy Commands, rather than a missing or unhealthy volume.
I have not detached/recreated the volume because Railway reports it as healthy and it contains persistent data.
Happy to provide deployment IDs if useful.
Status changed to Awaiting User Response Railway • about 1 month ago
a month ago
Hey! Thanks for the detailed report. This was caused by a recent change on our side, and we've now rolled it back. Deploys with a volume and a pre-deploy command should go through again, so please redeploy and your volume will be mounted as usual. Your volume and its data weren't affected.
If you detached your volume or deployed without it while this was happening, reply here before redeploying and we'll help make sure nothing written since then gets lost.
Sorry for the disruption, and let us know if anything still fails.
a month ago
Thanks, both. womega's workaround did it for me too: I removed the pre-deploy command and put the migration in the start command (migrate && start), and the next deploy went through with the volume mounted normally. I never detached the volume, so nothing to recover on my side. Good to hear it's rolled back; I'll move the migration back to a pre-deploy command.
Status changed to Awaiting Railway Response Railway • about 1 month ago
Status changed to Solved Railway • about 1 month ago
