We run four production Postgres databases on Railway (Pro workspace) and are

about to consolidate them into one. Today a Postgres failure affects one

customer; afterwards it would affect all of them, so we want replication and

failover in place before the migration rather than after.

Before asking, we checked what we could from the public GraphQL API:

• enablePitrForHaCluster / disablePitrForHaCluster exist
• pitrHaClusterReplicationHealth(environmentId, rootServiceId) exists
• cancelPitrHaWorkflow / pitrHaWorkflowProgress exist

So "Postgres HA cluster" is clearly a concept in the platform, but we can't

find any mutation that CREATES one, and templateSearch returns nothing for

"postgres ha" or "postgres high availability". Our existing Postgres services

report no replicas and pitrHaClusterReplicationHealth does not resolve for them.

Questions:

1. Is Postgres high availability (streaming replica + failover) available on the

   Pro plan? If so, how is it provisioned — dashboard flow, template, or via

   support?

2. If it is not self-serve, is it on the roadmap, or is it an Enterprise

   feature?

3. We already have WAL archiving / PITR enabled and healthy on all four

   databases (archive_mode=on, zero archive failures). Can PITR restore into a

   NEW service, and roughly how long does a restore take for a ~1 GB database?

   We're trying to decide whether a fast, proven restore is an acceptable

   substitute for failover at our scale.

4. To be explicit about something we want to avoid getting wrong: the Pro plan

   limit shows deploys.replicas = 42. We're assuming that is stateless service

   replicas and must NOT be used to scale a Postgres service backed by a single

   volume. Please confirm.

Happy to share project IDs privately if that helps.