Struct ServiceLifecycle 

pub struct ServiceLifecycle { /* private fields */ }

Shared process lifecycle and health-probe owner.

Implementations

impl ServiceLifecycle

pub fn new(config: ServiceLifecycleConfig) -> Self

Creates a new ServiceLifecycle.

pub fn from_env(service_name: impl Into<Arc<str>>) -> RuntimeResult<Self>

Creates a value from env.

pub fn config(&self) -> &ServiceLifecycleConfig

Returns the config.

pub fn set_observer( &self, observer: Arc<dyn ServiceLifecycleObserver>, ) -> RuntimeResult<()>

Binds one observer before the first state transition and emits Starting.

Errors

Returns an error if an observer is already installed or the lifecycle has left Starting. Installation is serialized with state transitions.

pub async fn start( &self, service_context: &ChildServiceContext, ) -> RuntimeResult<()>

Starts the progress heartbeat and optional HTTP probe server under a dedicated component task group of service_context.

Errors

Returns an error when called twice, when the listener cannot bind, or when either owned service task cannot be registered. A failed start cancels and awaits every task registered by that attempt before making the lifecycle retryable.

pub fn state(&self) -> ServiceLifecycleState

Returns the state.

pub fn probe_local_addr(&self) -> Option<SocketAddr>

Returns the probe local addr.

pub fn is_ready(&self) -> bool

Returns whether ready.

pub fn set_dependency_readiness(&self, readiness: DependencyReadiness)

Publishes the aggregate dependency decision without changing lifecycle state. A healthy dependency cannot undo maintenance, drain, failure, or stop.

pub fn dependency_readiness(&self) -> DependencyReadiness

Returns the latest aggregate dependency decision independently of lifecycle state.

pub fn handle_critical_failure( &self, failure: &CriticalFailure, recovery: CriticalFailureRecovery, )

Applies the handling policy for one critical task failure.

Dependency readiness is revoked first, because a process whose critical work failed must stop advertising itself as ready even when it keeps making progress. The remaining effects are the caller’s policy: this method never decides on its own whether the process fails or asks for an ordered shutdown.

pub fn spawn_critical_failure_monitor( &self, owner: &TaskSpawner, failures: &CriticalFailureState, recovery: CriticalFailureRecovery, ) -> RuntimeResult<TaskId>

Spawns a monitor that applies recovery when a critical failure is pending.

owner must be an owner outside the group of the monitored tasks, because a poisoned group can neither spawn nor run its own monitor. The monitor takes each pending failure before handling it, so repeated failures are handled in sequence rather than replayed.

Errors

Returns an error when owner is shutting down or closed.

pub fn is_live(&self) -> bool

Returns whether live.

pub fn record_progress(&self)

Records progress.

pub fn mark_ready(&self) -> RuntimeResult<()>

Marks startup dependencies ready without allowing a drain request to be reversed.

Errors

Returns an error when shutdown or failure already began.

pub fn suspend_readiness_for_maintenance(&self) -> RuntimeResult<()>

Temporarily removes this process from readiness during a reversible maintenance operation.

This does not alter the process lifecycle state. A real shutdown request remains irreversible and prevents a later readiness restore.

Errors

Returns an error unless the process is ready and shutdown has not begun.

pub fn restore_readiness_after_maintenance(&self) -> RuntimeResult<()>

Restores readiness after a reversible maintenance operation.

Errors

Returns an error when shutdown, failure, or stop has begun.

pub fn mark_failed(&self)

Executes mark failed.

pub fn mark_stopped(&self)

Executes mark stopped.

pub fn request_shutdown(&self, reason: ShutdownReason) -> ShutdownRequest

Records the first shutdown request and returns its immutable deadline.

Repeated pre-stop or signal delivery is idempotent and cannot extend the budget.

pub fn shutdown_request(&self) -> Option<ShutdownRequest>

Shuts down request.

pub async fn wait_for_shutdown(&self) -> ShutdownRequest

Returns the wait for shutdown.

pub async fn wait_for_shutdown_signal(&self) -> RuntimeResult<ShutdownRequest>

Waits for pre-stop or an operating-system termination signal.

Errors

Returns the platform signal registration error without inventing a successful signal.

Trait Implementations

impl Clone for ServiceLifecycle

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 ServiceLifecycle

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

Formats the value using the given formatter.