Struct ConcurrencyLimit 

pub struct ConcurrencyLimit { /* private fields */ }

Allow at most max requests in flight, answer the rest with 503 and Retry-After

By default an excess request is rejected at once. ConcurrencyLimit::queue_for lets it wait a short while for a free slot first, which smooths bursts without building an unbounded backlog.

Implementations

impl ConcurrencyLimit

pub fn new(max: usize) -> Self

pub fn queue_for(self, wait: Duration) -> Self

Let an excess request wait up to wait for a slot before it is rejected

pub fn retry_after(self, seconds: u64) -> Self

Seconds advertised in Retry-After (default 1)

pub fn available_slots(&self) -> usize

Requests that currently hold a slot is max - available_slots()

Trait Implementations

impl Middleware for ConcurrencyLimit

fn handle<'life0, 'async_trait>( &'life0 self, req: Request, next: Next, ) -> Pin<Box<dyn Future<Output = Result<Response>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,