Hello Railway Support,

We have a Redis service in a staging environment that is no longer referenced by any of our application services. We attempted to suspend it reversibly while preserving its attached persistent volume.

Context:

• Railway CLI version: 5.26.1
• Environment: staging
• Service type: Redis
• Persistent volume attached and Ready
• Previous deployment metadata reported the regional configuration as sfo=1
• Previous deployment status: SUCCESS

Command executed:

railway scale --service Redis --environment staging sfo=0

Observed result:

• The command exited successfully with exit code 0.
• Railway created a new successful deployment.
• The service never reached zero replicas.
• The resulting deployment continued with desired replicas set to 1.
• One instance remained RUNNING.
• The persistent volume remained attached and Ready.
• No validation error or warning was returned.
• The generated deployment timestamp was 2026-07-15T12:58:07Z.
• The service is currently healthy with one desired replica and one running instance.

We did not repeat the scale-to-zero command and did not delete or detach the volume.

Could you please confirm:

1. Is scaling a service with an attached persistent volume to zero replicas supported?
2. Does Railway enforce a minimum of one replica for database services or services with volumes?
3. Is sfo still a supported region alias for scale commands, or should a different current region identifier be used?
4. Why did the command return success and generate a deployment while preserving one desired replica?
5. What is the recommended reversible procedure for stopping this Redis service while preserving its service configuration and persistent volume?

We can provide project, service, and deployment identifiers privately if needed.

Please do not make any changes to the service on our behalf without our explicit confirmation.

Best regards,