Struct ClusterManager

pub struct ClusterManager {
    pub state: Arc<RwLock<SharedState>>,
    pub nodes: Arc<RwLock<HashMap<Arc<str>, NodeInfo>>>,
}

Central manager for cluster operations and node tracking.

The ClusterManager is responsible for maintaining the state of the entire cluster, including tracking which nodes are currently active, managing node registration and removal, and providing information about the cluster’s current composition.

It uses thread-safe data structures to allow concurrent access from multiple parts of the application, particularly important in a distributed system where node changes can happen at any time.

Fields

• state - Shared state that includes information about the current node and cluster
• nodes - Thread-safe map of all known nodes in the cluster, indexed by their address

Fields

state: Arc<RwLock<SharedState>>

Shared state containing information about the current node and overall cluster

nodes: Arc<RwLock<HashMap<Arc<str>, NodeInfo>>>

Thread-safe map of all nodes in the cluster, keyed by node address

Implementations

impl ClusterManager

pub fn new(state: Arc<RwLock<SharedState>>) -> Self

Creates a new instance of the ClusterManager.

Initializes a new cluster manager with the provided shared state and an empty nodes map. This is typically called during application startup to establish the cluster management subsystem.

Arguments

• state - Shared state containing information about the current node

Returns

A new ClusterManager instance ready to track and manage cluster nodes

pub async fn register_node(&self, node: NodeInfo)

Registers a new node in the cluster.

This method adds a node to the cluster’s node registry if it doesn’t already exist. After registration, it updates the shared state with the new cluster size. The method uses colorized output to make debugging in the console easier.

Arguments

• node - Information about the node to register

Side Effects

• Updates the internal nodes map if the node is new
• Updates the cluster size in the shared state
• Prints diagnostic information to console

pub async fn remove_node(&self, node_uid: Arc<str>)

Removes a node from the cluster.

This method removes a node from the cluster’s node registry if it exists. After removal, it updates the shared state with the new cluster size. The method includes debug logging to track node removal operations.

Arguments

• node_uid - Unique identifier of the node to remove, typically its address

Side Effects

• Updates the internal nodes map by removing the specified node
• Updates the cluster size in the shared state
• Logs diagnostic information about the removal operation

pub async fn get_nodes(&self) -> Vec<NodeInfo>

Retrieves a list of all known nodes in the cluster.

This method provides a snapshot of all the nodes currently registered in the cluster manager. It’s useful for operations that need to iterate over all nodes or display cluster status information.

Returns

A vector containing information about all known nodes in the cluster

pub async fn get_nodes_and_self(&self) -> Vec<NodeInfo>

Retrieves a list of all known nodes plus the current node.

This method provides a complete view of the cluster including the current node. It’s particularly useful for operations that need a complete picture of the cluster, such as leader election or quorum calculations.

Returns

A vector containing information about all nodes in the cluster, including the current node

Side Effects

• Prints diagnostic information about the current node and known nodes to console

pub async fn is_node_alive(&self, node_uid: Arc<str>) -> bool

Checks if a specific node is currently active in the cluster.

This method determines if a node is still considered active within the cluster by checking if it exists in the nodes registry. It’s useful for operations that need to verify a node’s presence before attempting to communicate with it.

Arguments

• node_uid - Unique identifier of the node to check, typically its address

Returns

true if the node is active in the cluster, false otherwise

impl ClusterManager

pub async fn discover_peers( &self, config: &ServerConfig, my_port: u16, ) -> Result<()>

Discovers and connects to peer nodes in the cluster

Arguments

• config - Server configuration containing instance information
• my_port - Current node’s port to avoid self-connection

Returns

Result indicating success or failure of the discovery process

pub(crate) async fn connect_to_peer( &self, client: &Client, node_address: &str, ) -> Result<()>

Attempts to establish a connection with a peer node

Arguments

• client - HTTP client for making requests
• node_address - Address of the peer node

Returns

Result indicating success or failure of the connection attempt

Trait Implementations

impl Debug for ClusterManager

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.