Struct DistributedConfig 

pub struct DistributedConfig {
    pub role: NodeRole,
    pub bind_addr: SocketAddr,
    pub coordinator_addr: Option<SocketAddr>,
    pub expect_workers: usize,
    pub heartbeat_interval_ms: u64,
    pub heartbeat_timeout_ms: u64,
    pub node_id: String,
}

Configuration for distributed training.

Fields

role: NodeRole

Role of this node

bind_addr: SocketAddr

Address to bind (coordinator) or connect to (worker)

coordinator_addr: Option<SocketAddr>

Coordinator address (workers only; coordinators use bind_addr)

expect_workers: usize

Expected number of workers (coordinator only, for barrier)

heartbeat_interval_ms: u64

Heartbeat interval in milliseconds

heartbeat_timeout_ms: u64

Heartbeat timeout in milliseconds (detect failure after this)

node_id: String

Node identifier (auto-assigned from hostname + pid)

Implementations

impl DistributedConfig

pub fn coordinator(bind_addr: SocketAddr, expect_workers: usize) -> Self

Create a coordinator config.

Arguments

• bind_addr - Address to listen on (e.g., 0.0.0.0:9000)
• expect_workers - Total worker count (including coordinator’s own GPUs)

pub fn worker(coordinator_addr: SocketAddr) -> Self

Create a worker config.

Arguments

• coordinator_addr - Address of the coordinator (e.g., intel:9000)

pub fn is_coordinator(&self) -> bool

Check if this node is a coordinator

Trait Implementations

impl Clone for DistributedConfig

fn clone(&self) -> DistributedConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for DistributedConfig

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

Formats the value using the given formatter.

impl Default for DistributedConfig

fn default() -> Self

Returns the “default value” for a type.