Struct HttpHealthCheck

pub struct HttpHealthCheck {
    pub consecutive_success: usize,
    pub consecutive_failure: usize,
    pub peer_template: HttpPeer,
    pub reuse_connection: bool,
    pub req: RequestHeader,
    pub validator: Option<Box<dyn Fn(&ResponseHeader) -> Result<()> + Send + Sync>>,
    pub port_override: Option<u16>,
    pub health_changed_callback: Option<HealthObserveCallback>,
    pub backend_summary_callback: Option<BackendSummary>,
    /* private fields */
}

HTTP health check

This health check checks if it can receive the expected HTTP(s) response from the given backend.

Fields

consecutive_success: usize

Number of successful checks to flip from unhealthy to healthy.

consecutive_failure: usize

Number of failed checks to flip from healthy to unhealthy.

peer_template: HttpPeer

How to connect to the backend.

This field defines settings like the connect timeout and src IP to bind. The SocketAddr of peer_template is just a placeholder which will be replaced by the actual address of the backend when the health check runs.

Set the scheme field to use HTTPs.

reuse_connection: bool

Whether the underlying TCP/TLS connection can be reused across checks.

• false will make sure that every health check goes through TCP (and TLS) handshakes. Established connections sometimes hide the issue of firewalls and L4 LB.
• true will try to reuse connections across checks, this is the more efficient and fast way to perform health checks.

req: RequestHeader

The request header to send to the backend

validator: Option<Box<dyn Fn(&ResponseHeader) -> Result<()> + Send + Sync>>

Optional field to define how to validate the response from the server.

If not set, any response with a 200 OK is considered a successful check.

port_override: Option<u16>

Sometimes the health check endpoint lives one a different port than the actual backend. Setting this option allows the health check to perform on the given port of the backend IP.

health_changed_callback: Option<HealthObserveCallback>

A callback that is invoked when the healthy status changes for a Backend.

backend_summary_callback: Option<BackendSummary>

An optional callback for backend summary reporting.

Implementations

impl HttpHealthCheck

pub fn new(host: &str, tls: bool) -> Self

Create a new HttpHealthCheck with the following default settings

• connect timeout: 1 second
• read timeout: 1 second
• req: a GET to the / of the given host name
• consecutive_success: 1
• consecutive_failure: 1
• reuse_connection: false
• validator: None, any 200 response is considered successful

pub fn set_connector(&mut self, connector: HttpConnector)

Replace the internal http connector with the given HttpConnector

pub fn set_backend_summary<F>(&mut self, callback: F)

where F: Fn(&Backend) -> String + Send + Sync + 'static,

Trait Implementations

impl HealthCheck for HttpHealthCheck

fn health_threshold(&self, success: bool) -> usize

This function defines how many consecutive checks should flip the health of a backend.

fn check<'life0, 'life1, 'async_trait>( &'life0 self, target: &'life1 Backend, ) -> Pin<Box<dyn Future<Output = Result<()>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Check the given backend.

fn health_status_change<'life0, 'life1, 'async_trait>( &'life0 self, target: &'life1 Backend, healthy: bool, ) -> Pin<Box<dyn Future<Output = ()> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Called when the health changes for a Backend.

fn backend_summary(&self, target: &Backend) -> String

Called when a detailed Backend summary is needed.