Skip to content
Inspect a stopped Railway-built cron image without restarting the worker
stehlc
OP

a month ago

Could Railway confirm the supported recovery options for a stopped cron deployment built from a CLI upload?

I found this related answer from Railway's support bot: https://station.railway.com/questions/export-or-inspect-the-exact-source-image-dde7a9d5 . It says source/image export is unavailable and file inspection requires a running container. I would like to confirm whether that limitation applies here, or whether the owning workspace has any supported escalation or recovery path.

The deployment metadata has an image digest but no Git revision, source repository URL, or pullable image reference. SSH cannot inspect the stopped container. The deployment-snapshot API exposes configuration rather than a source archive. No other service in the affected project runs the exact same image. We have not restarted or redeployed the worker.

We need to inspect routing/configuration before separating one workload from a shared job. Restarting its normal command could repeat external submissions. Please do not restart the worker or execute its normal command as part of this request.

If immutable-image export, original source archive/revision mapping and non-executing inspection are all unavailable, please confirm that limitation so we can plan recovery from the original build machine instead.

This is a platform-capability question, so Service is N/A. The affected service is not available in this account's service selector. Client details, infrastructure identifiers, logs, credentials and customer data are omitted from this public thread. If staff need exact deployment identifiers, what private handoff route should the owning workspace use? An earlier email request was directed to Central Station.

Solved

3 Replies

Railway
BOT

a month ago

That limitation applies here too. We don't provide exports of the built image, the uploaded source archive, or a mapping from a deployment back to its original source. There is also no way to inspect a stopped deployment's files without running it, and no recovery or escalation route that does this on the workspace's behalf.

You can only get files off a running container, through the service's Console and Files panel, railway ssh, or scp/sftp. If the deployment is still within your plan's image retention window, Rollback can restore its image, but that starts the deployment with its normal start command, which is what you need to avoid.

You don't need to send deployment identifiers, because they wouldn't change any of this. Recovering from the original build machine is the right plan.


Status changed to Awaiting User Response Railway • about 1 month ago


Railway

That limitation applies here too. We don't provide exports of the built image, the uploaded source archive, or a mapping from a deployment back to its original source. There is also no way to inspect a stopped deployment's files without running it, and no recovery or escalation route that does this on the workspace's behalf. You can only get files off a running container, through the service's Console and Files panel, `railway ssh`, or `scp`/`sftp`. If the deployment is still within your plan's [image retention window](https://docs.railway.com/pricing/plans#image-retention-policy), Rollback can restore its image, but that starts the deployment with its normal start command, which is what you need to avoid. You don't need to send deployment identifiers, because they wouldn't change any of this. Recovering from the original build machine is the right plan.

stehlc
OP

a month ago

Thanks. Please leave this unresolved and have a Railway team member confirm these platform limitations, if possible. We will pursue the original build machine; no restart, rollback, redeployment, start-command change, or resource creation is authorized by this request. The remaining question is whether there is any supported read-only recovery path that avoids executing the worker.


Status changed to Awaiting Railway Response Railway • about 1 month ago


Railway
BOT

a month ago

We can confirm the limitations in our earlier reply. There is no supported read-only path for a stopped deployment: we don't export the built image, hand back the uploaded source archive, or map a deployment back to its original source, and there's no way to mount or browse a stopped deployment's filesystem without running it.

Files can only be read from a running container, through the service's Console and Files panel, railway ssh, or scp/sftp. Rollback can bring back an image that is still inside your plan's image retention window, but it starts that deployment with its normal start command, so it doesn't meet your no-execution requirement.

Deployment identifiers wouldn't change any of this, so you don't need to send them. Recovering from the original build machine is the right path.


Status changed to Awaiting User Response Railway • about 1 month ago


Railway
BOT

24 days ago

This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!

Status changed to Solved Railway • 24 days ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...