Struct PeerGovernor 

pub struct PeerGovernor { /* private fields */ }

Per-peer policy + health state owner.

Snapshot/restore: full state is captured in FrameworkSnapshot (work). Pre-Stage-5, restore rebuilds the governor from scratch - blocklist + allowlist settings on NodeConfig are re-applied at construction time.

Implementations

impl PeerGovernor

pub fn new() -> Self

Construct a fresh governor with no blocklist, no allowlist, and the default failure threshold.

pub fn with_failure_threshold(self, threshold: u32) -> Self

Configure the failure threshold (consecutive failures before a peer is marked down). Default DEFAULT_FAILURE_THRESHOLD.

pub fn block(&mut self, peer: PeerId)

Add a peer to the blocklist. Subsequent inbound envelopes from this peer are dropped; subsequent outbound sends fail.

pub fn unblock(&mut self, peer: PeerId)

Remove a peer from the blocklist.

pub fn set_allowlist(&mut self, allowlist: Option<HashSet<PeerId>>)

Set (or clear) the allowlist. When Some, only peers in the allowlist may communicate; everyone else is denied. None means “open” (default).

pub fn peer_health(&self, peer: PeerId) -> Option<PeerHealth>

Per-peer health snapshot read; returns None if no success/failure has ever been recorded for peer.

pub fn is_down(&self, peer: PeerId) -> bool

True when peer is currently marked down.

pub fn check_inbound(&self, peer: PeerId) -> Decision

Inbound consultation - called from the engine’s Phase 1 envelope router before any slot is written.

pub fn check_outbound( &self, peer: PeerId, backoff: &BackoffTable, now_ns: u64, ) -> Decision

Outbound consultation - called from the compiler-inserted PeerHealthGate syscall op upstream of every wire::Send. Returns Deny(Cooldown { retry_ns }) when the peer is in failure cooldown, prompting the gate to reschedule.

pub fn record_success( &mut self, peer: PeerId, now_ns: u64, ) -> LifecycleTransition

Record a successful exchange with peer at now_ns. Resets the consecutive-failure counter; clears down. Returns the lifecycle transition the engine should surface as an EngineStep::PeerUp.

pub fn record_failure( &mut self, peer: PeerId, now_ns: u64, ) -> LifecycleTransition

Record a failure for peer at now_ns. Returns WentDown when the failure pushes the streak across the threshold.

pub fn tracked_peers(&self) -> usize

Number of distinct peers with health state recorded.

pub fn blocklist(&self) -> &HashSet<PeerId>

Read-only view of the blocklist. Used by snapshot capture.

pub fn allowlist(&self) -> Option<&HashSet<PeerId>>

Read-only view of the allowlist.

pub fn iter_health(&self) -> impl Iterator<Item = (PeerId, PeerHealth)> + '_

Iterate (PeerId, PeerHealth) entries for snapshot capture.

pub fn failure_threshold(&self) -> u32

Current failure threshold (consecutive failures to mark a peer down). Used by snapshot capture.

pub fn restore_health(&mut self, peer: PeerId, health: PeerHealth)

Replace a peer’s health state directly. Used by Node::restore to re-seed health from a snapshot entry.

pub fn set_failure_threshold(&mut self, threshold: u32)

Overwrite the failure threshold. Used by Node::restore.

Trait Implementations

impl Default for PeerGovernor

fn default() -> Self

Returns the “default value” for a type.