Trait tor_hsservice::publish::backoff::BackoffSchedule

pub(super) trait BackoffSchedule {
    // Required methods
    fn max_retries(&self) -> Option<usize>;
    fn overall_timeout(&self) -> Option<Duration>;
    fn single_attempt_timeout(&self) -> Option<Duration>;
    fn next_delay<E: RetriableError>(&mut self, error: &E) -> Option<Duration>;
}

A trait that specifies the parameters for retrying a fallible operation.

Required Methods

fn max_retries(&self) -> Option<usize>

The maximum number of retries.

A return value of None indicates is no upper limit for the number of retries, and that the operation should be retried until BackoffSchedule::overall_timeout time elapses (or indefinitely, if BackoffSchedule::overall_timeout returns None).

fn overall_timeout(&self) -> Option<Duration>

The total amount of time allowed for the retriable operation.

A return value of None indicates the operation should be retried until BackoffSchedule::max_retries number of retries are exceeded (or indefinitely, if BackoffSchedule::max_retries returns None).

fn single_attempt_timeout(&self) -> Option<Duration>

The total amount of time allowed for a single operation.

fn next_delay<E: RetriableError>(&mut self, error: &E) -> Option<Duration>

Return the delay to introduce before the next retry.

The error parameter contains the error returned by the fallible operation. This enables implementors to (optionally) implement adaptive backoff. For example, if the operation is sending an HTTP request, and the error is a 429 (Too Many Requests) HTTP response with a Retry-After header, the implementor can implement a backoff schedule where the next retry is delayed by the value specified in the Retry-After header.

Object Safety

This trait is not object safe.

Implementors

impl<M: Mockable> BackoffSchedule for PublisherBackoffSchedule<M>