Hi, we've been cleaning up our volumes and accidentally deleted a volume for a database that we were still using. Although this was DEV data, we didn't manage to re-attach the volume and would like to know for future reference what we did wrong.

What we attempted

We attaching the volume to an existing service - it was successful:

railway volume attach -v poetic-volume

but this command has no -s or --service parameter - so it could only be attached to the random service the command suggested in our case our-service

We had the following shown during railway volume list:

Volume: poetic-volume
Attached to: our-service
Mount path: /var/lib/postgresql/data
Storage used: 1689MB/50000MB
Status: Ready
Deletes on: Jul 23 2026 8:23 PM +02:00

Volume was attached, but somehow still in soft-delete mode and not browseable.

railway volume browse -v poetic-volume

and variations of railway volume files all errored. The dashboard also didn't show the volume attached to our randomly selected our-service service (despite the cli response saying otherwise). We tried both web ui redeploy and CLI command redeploy: railway redeploy --service our-service - but the volume was still inaccessible (it also wasn't visible on the web ui near this service).

Oddly enough detaching also didn't work.

railway volume attach -v poetic-volume showed:

Volume "poetic-volume" detached from service "our-service"

but subsesquent railway volume list still shows it ghost-attached to our-service.

No urgency as we managed to mitigate (lost dev data) but for future references we'd like to know how to mitigate the lack of -s (service picker) param on attach, or what is the suggested way of retrieving a soft-deleted volume (or reach it's data for manual recovery).

Thank you in advance!