5 hours ago
I need to restore only my existing autostack-landing service (production), serving info.gestiondeautos.com. The service was intentionally stopped on November 17; all other services should remain stopped.
The original deployment was uploaded with the Railway CLI on April 1, 2026. The API still reports canRollback=true and canRedeploy=true. However, rollback attempts on November 17 and November 19 fail before the application can start. The latest attempt, on November 19 at 15:07:18 UTC, failed at 15:12:57 UTC during Deploy > Create container.
The Railway platform error is: “The registry could not be reached to pull the image” from the production-us-east4-eqdc4a.railway-registry.com registry. No application logs are generated. The landing currently returns Railway’s 404 “Application not found”.
Redeploying the same original deployment with usePreviousImageTag=false also failed in BUILD_IMAGE with “Deployment failed” and no build logs. The service is linked to this request so the Railway team can identify the relevant deployments.
Could the Railway team check whether the original image is still retained and restore or re-publish it so this service can start? If the image is unavailable, can Railway rebuild the original CLI-uploaded source for that deployment? Please preserve the existing service and custom-domain configuration and keep the other services stopped.
The local folder we recovered contains an earlier prototype, not the Spanish landing that was actually deployed, so uploading that prototype would not restore the existing website. I am asking for help recovering the original deployment, not for a redesign.
Thank you.
1 Reply
Status changed to Awaiting Railway Response Railway • about 5 hours ago
4 hours ago
Your November 19 rollback reached the container step, then the image pull from our registry was refused, so the container never started. That is why no application logs were produced and the domain shows the 404 page. Nothing in your service settings, variables or custom-domain configuration causes that pull to fail, and there is no change on your side that would make it succeed.
Redeploying with usePreviousImageTag=false takes a different path: it rebuilds from that deployment's originally uploaded source instead of pulling the stored image, which is why that attempt stopped at the build step rather than the pull.
Rebuilding, re-publishing or handing back a deployment's image or uploaded source as custom work isn't something we do, so the rollback and redeploy you already tried are the restore paths for that exact version. Both act only on the autostack-landing service, so your other services stay stopped either way.
Status changed to Awaiting User Response Railway • about 4 hours ago