Struct BranchMonitor 

pub struct BranchMonitor { /* private fields */ }

Implementations

impl BranchMonitor

pub fn new(node_id: &str) -> Self

pub fn with_sink(self, sink: SharedSink) -> Self

Attach an observation sink (JSON Lines, OTel, database adapter, …).

pub fn with_service(self, service: impl Into<String>) -> Self

Identify the service/component this monitor runs in, for the observations it emits.

pub fn with_service_version(self, version: impl Into<String>) -> Self

Identify the software version generating observations, distinct from the deployment slot. Lets an analyst compare “predictions from build v1” against “observations generated by v1”.

pub fn with_deployment(self, deployment: impl Into<String>) -> Self

Identify the deployment/environment (e.g. prod-us-east-1).

pub fn with_build_ref(self, build_ref: impl Into<String>) -> Self

A stable reference to the compiled reliability artifact/build that produced this service, so observations can be traced back to the model that predicted them without embedding the whole artifact.

pub fn record_failure_observation(&self, observation: ReliabilityObservation)

Record an immutable failure observation for offline analysis. Lightweight: does not run statistics, query databases, or call AI.

pub fn with_sla(node_id: &str, sla_tracker: Arc<Mutex<SlaTracker>>) -> Self

pub fn with_chaos(node_id: &str, chaos: Arc<Mutex<ChaosController>>) -> Self

pub fn record_branch(&mut self, outcome: &str, declared_probability: f64)

pub fn record_failure( &mut self, operation_id: &str, error: &dyn Error, declared_probability: Option<f64>, )

pub fn record_success( &mut self, operation_id: &str, declared_probability: Option<f64>, )

Record that operation_id completed without failing. Must be called on the same "{operation_id}.failure" SLA key [record_failure] uses, with occurred = false, or that key’s rolling window only ever sees failures (every entry record_failure ever pushes) and its observed frequency is permanently 1.0 — which made [record_failure]’s anomaly check fire unconditionally for any operation that failed at least a handful of times over its lifetime (sla::MIN_OBSERVATIONS), regardless of its actual overall failure rate. Generated code calls this from the operation’s Ok arm whenever it calls record_failure from the matching Err arm (i.e. whenever onFailureProbabilitySource is declared and resolves) — see codegen/rust.rs’s Ok(_result) => arm.

pub fn flush(&self)

Trait Implementations

impl Drop for BranchMonitor

fn drop(&mut self)

Executes the destructor for this type.

fn pin_drop(self: Pin<&mut Self>)

🔬This is a nightly-only experimental API. (pin_ergonomics)

Execute the destructor for this type, but different to Drop::drop, it requires self to be pinned.