Struct DistributedCoordinator 

pub struct DistributedCoordinator { /* private fields */ }

In-memory multi-node coordinator.

Manages cluster topology, routes incoming requests to healthy nodes via consistent hashing, and exposes cluster health information.

Note: This implementation is intentionally in-memory only — no actual TCP connections are established. It is designed for unit testing and single-process simulation. Production deployments would wrap this with a gRPC/HTTP gossip layer.

Implementations

impl DistributedCoordinator

pub fn new(config: CoordinatorConfig) -> Self

Create a new coordinator with the given configuration.

Does not automatically register self — call register_self to add this node to the ring.

pub fn register_self(&mut self)

Register this node in the local registry so it participates in routing.

pub fn add_peer(&mut self, addr: &str, node_id: &str)

Add a peer to the registry as a healthy node with zero load.

addr is the peer’s "host:port" bind address. node_id is the peer’s unique identifier.

pub fn route(&self, request_key: &str) -> Option<String>

Route request_key to a healthy node and return its address.

Returns None if no healthy nodes are available.

pub fn cluster_size(&self) -> usize

Total number of nodes registered in the cluster (healthy + unhealthy).

pub fn healthy_count(&self) -> usize

Number of nodes currently marked as healthy.

pub fn topology_summary(&self) -> String

A human-readable summary of current cluster topology.

Format (not stable across versions):

cluster[nodes=3 healthy=2 vnodes=450 self=node-0]

pub fn registry(&self) -> &NodeRegistry

Access the underlying registry (read-only).

pub fn config(&self) -> &CoordinatorConfig

Access the coordinator config.

pub fn set_peer_health(&mut self, node_id: &str, healthy: bool)

Mark a peer node as healthy or unhealthy.

pub fn update_peer_load(&mut self, node_id: &str, load: f32)

Update a peer’s reported load factor.