Struct NetworkHealthMonitor 

pub struct NetworkHealthMonitor { /* private fields */ }

Network health monitor

Implementations

impl NetworkHealthMonitor

pub fn new(cluster: Arc<ClusterState>, config: HealthConfig) -> Self

Create new health monitor

pub fn with_runtime_profile( cluster: Arc<ClusterState>, profile: &RuntimeProfile, ) -> Self

Create a monitor with thresholds derived from runtime auto-detection.

pub fn config(&self) -> HealthConfig

Read the current configuration lock-free.

pub fn register_tcp_probe_target( &self, node_id: impl Into<String>, target: SocketAddr, )

Register a node-specific TCP probe target.

pub fn clear_tcp_probe_target(&self, node_id: &str) -> bool

Remove a TCP probe target for a node.

pub fn check_all(&self)

Run health check on all nodes

pub fn get_node_health(&self, node_id: &str) -> Option<HealthCheckResult>

Get health report for a specific node

pub fn get_all_health(&self) -> Vec<HealthCheckResult>

Get health report for all nodes

pub fn healthy_node_count(&self) -> usize

Get the number of nodes whose latest health sample is healthy.

pub fn needs_quantization_fallback(&self, node_id: &str) -> bool

Check if node needs quantization fallback

OPTIMIZATION: Accumulates delivery ratio and latency across recent checks in a single pass without allocating intermediate candidate vectors or performing multiple iterator traversals.

pub fn get_health_summary(&self) -> String

Get cluster-wide health summary

OPTIMIZATION: Consolidates node health counts into a single-pass traversal over recent_checks, eliminating three redundant iterator traversals and repeated back() calls per node.

pub fn start_periodic_checks(&self)

Start periodic health checks in background

pub fn reconfigure_from_system_profile(&self, profile: &SystemProfile)

Re-tune health thresholds from a system profile detected at runtime.

This is the main integration point with SystemProfileWatcher: when hardware changes (GPU hot-plug, CPU scaling), call this to keep the health monitor’s expectations aligned with reality.

pub fn subscribe_to_watcher( self: &Arc<Self>, watcher: &SystemProfileWatcher, ) -> JoinHandle<()>

Subscribe to a SystemProfileWatcher’s broadcast channel and automatically re-tune whenever hardware changes are detected.

Returns a JoinHandle that can be cancelled by dropping it.

Trait Implementations

impl Clone for NetworkHealthMonitor

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for NetworkHealthMonitor

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

Formats the value using the given formatter.