Prevent duplicate node identifiers in a cluster #124
Description
Each server (node) in a cluster uses an identifier that is expected to be unique within that cluster.
As a result of OF-3181, cluster nodes can now be configured with a static, human-defined identifier. Previously, this identifier was always generated as a random UUID, which implicitly guaranteed uniqueness.
Because this change removes an important source of randomness, additional guardrails should be introduced to prevent misconfiguration. In particular, the system should detect and reject cluster states in which multiple nodes share the same identifier. For example, a server should be prevented from joining a cluster if it detects that its configured identifier is already in use by another node in that cluster.