Struct RegressionHistoryPolicy 

pub struct RegressionHistoryPolicy { /* private fields */ }

Adaptive SLO policy backed by per-target blocked-ratio history.

Uses bounded threshold shifts around default class SLOs. History can be persisted to JSON for process restarts and operator inspection.

Implementations

impl RegressionHistoryPolicy

pub const fn new() -> Self

Create an in-memory adaptive policy with no persistence.

Example

use stygian_charon::RegressionHistoryPolicy;

let policy = RegressionHistoryPolicy::new();
assert!(policy.tracked_target_count() == 0);

pub fn with_json_store( path: impl AsRef<Path>, ) -> Result<Self, AdaptivePolicyError>

Create an adaptive policy that persists history to a JSON file.

If the file exists, it is loaded on startup.

Errors

Returns AdaptivePolicyError when loading the store fails.

Example

use stygian_charon::RegressionHistoryPolicy;

let _policy = RegressionHistoryPolicy::with_json_store("./charon-history.json")?;

pub fn tracked_target_count(&self) -> usize

Number of unique targets tracked by history.

pub fn observations_for_target(&self, target: &str) -> usize

Number of observations currently retained for one target.

Trait Implementations

impl AdaptiveSloPolicy for RegressionHistoryPolicy

fn select_slo( &self, target: &str, target_class: TargetClass, default: BlockedRatioSlo, ) -> BlockedRatioSlo

Select an adjusted SLO for a target using historical observations.

fn record_observation( &self, target: &str, target_class: TargetClass, blocked_ratio: f64, escalation_level: RequirementLevel, ) -> Result<(), AdaptivePolicyError>

Record a new blocked-ratio observation for a target.

impl Debug for RegressionHistoryPolicy

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

Formats the value using the given formatter.

impl Default for RegressionHistoryPolicy

fn default() -> Self

Returns the “default value” for a type.