24 days ago
Hello Railway team,
We have two retained volumes mounted at /data on API services with no active deployment. The workspace is now on an active Hobby subscription. Both volumes were observed Ready with no pending deletion on October 28.
The installed authenticated CLI volume-file command stops with: "Service api has no active deployment ... Deploy or restart the service before using file commands." No data download has completed.
We need one-time local copies of the existing files. The original application startup may run database migrations, so redeploying the old application is unsuitable. We have not detached, restored, wiped, or moved either volume. The dashboard says creating native backups requires Pro; our total monthly budget is USD 10, so we cannot upgrade to Pro.
I found the existing "Railway Observatory Volume Data Recovery" discussion (https://station.railway.com/questions/railway-observatory-volume-data-recovery-20428aeb), which says file commands need a running deployment. Could you confirm the safest supported procedure for an already-paid Hobby workspace?
- Is an offline read-only export or independent filesystem clone available?
- If a minimal recovery container must temporarily replace the application on the SAME existing service while keeping the existing volume attached, what exact source/build/start/pre-deploy and auto-deploy settings prevent the old application or its migrations from running? Does this avoid redeploying any other service?
- Can the original filesystem be mounted read-only for export? If not, please describe the supported precautions and limitations.
- How should that recovery deployment be stopped afterward while leaving the volume retained, without triggering the original application?
Please provide instructions and cost implications only. Do not modify services, volumes, or the subscription on our behalf. We can provide identifiers through an approved private channel if needed; none are included in this public question.
Thank you.
1 Reply
Status changed to Awaiting Railway Response Railway • 24 days ago
24 days ago
The volume file commands, railway volume browse, and sftp/scp all reach a volume through the service it is attached to, so they only work while that service has an active deployment. There is no offline export, separate filesystem clone, or read-only mount option for a volume. A volume is mounted read/write when the container starts, and it is not mounted during the build or the pre-deploy step.
Whatever deployment runs on that service gets the volume. That deployment uses whatever source, start command, and pre-deploy command the service has configured when it deploys. So the old application and its migrations only run if those settings still point at it. Any pre-deploy command still configured will run, but it cannot touch the volume. Only one deployment can be active and mounted on a service with a volume at a time.
Volume data only reaches your machine while something is running on that service. While a deployment runs, you pay normal per-minute compute for it, and storage keeps billing for the data on the volume as usual. Both rates are on the pricing page.
Status changed to Awaiting User Response sam-a • 24 days ago
17 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 • 17 days ago