Module aws_smithy_http_server::request::request_id

source ·
Available on crate feature request-id only.
Expand description

§Request IDs

aws-smithy-http-server provides the ServerRequestId.

§ServerRequestId

A ServerRequestId is an opaque random identifier generated by the server every time it receives a request. It uniquely identifies the request within that service instance. It can be used to collate all logs, events and data related to a single operation. Use ServerRequestIdProviderLayer::new to use ServerRequestId in your handler.

The ServerRequestId can be returned to the caller, who can in turn share the ServerRequestId to help the service owner in troubleshooting issues related to their usage of the service. Use ServerRequestIdProviderLayer::new_with_response_header to use ServerRequestId in your handler and add it to the response headers.

The ServerRequestId is not meant to be propagated to downstream dependencies of the service. You should rely on a distributed tracing implementation for correlation purposes (e.g. OpenTelemetry).

§Examples

Your handler can now optionally take as input a ServerRequestId.

ⓘ
pub async fn handler(
    _input: Input,
    server_request_id: ServerRequestId,
) -> Output {
    /* Use server_request_id */
    todo!()
}

let config = ServiceConfig::builder()
    // Generate a server request ID and add it to the response header.
    .layer(ServerRequestIdProviderLayer::new_with_response_header(HeaderName::from_static("x-request-id")))
    .build();
let app = Service::builder(config)
    .operation(handler)
    .build().unwrap();

let bind: std::net::SocketAddr = format!("{}:{}", args.address, args.port)
    .parse()
    .expect("unable to parse the server bind address and port");
let server = hyper::Server::bind(&bind).serve(app.into_make_service());

Structs§