7.1 KiB
title | menu | weight |
---|---|---|
Server | docs_basics | 150 |
The HTTP Server
The HttpServer type is responsible for serving http requests.
HttpServer
accepts an application factory as a parameter, and the application factory
must have Send
+ Sync
boundaries. More about that in the multi-threading section.
To bind to a specific socket address, bind()
must be used, and it may be
called multiple times. To bind ssl socket, bind_openssl()
or
bind_rustls()
should be used. To run the http server, use HttpServer::run()
method.
{{< include-example example="server" section="main" >}}
run()
method returns instance of Server
type. Methods of server type
could be used for managing http server
pause()
- Pause accepting incoming connectionsresume()
- Resume accepting incoming connectionsstop()
- Stop incoming connection processing, stop all workers and exit
Following example shows how to start http server in separate thread.
{{< include-example example="server" file="signals.rs" section="signals" >}}
Multi-threading
HttpServer
automatically starts a number of http workers, by default this number is
equal to number of logical CPUs in the system. This number can be overridden with the
HttpServer::workers()
method.
{{< include-example example="server" file="workers.rs" section="workers" >}}
Once the workers are created, they each receive a separate application instance to handle requests. Application state is not shared between the threads, and handlers are free to manipulate their copy of the state with no concurrency concerns.
Application state does not need to be
Send
orSync
, but application factory must beSend
+Sync
.
To share state between worker threads, use an Arc
. Special care should be taken once sharing and
synchronization are introduced. In many cases, performance costs are inadvertently introduced as a
result of locking the shared state for modifications.
In some cases these costs can be alleviated using more efficient locking strategies, for example using read/write locks instead of mutexes to achieve non-exclusive locking, but the most performant implementations often tend to be ones in which no locking occurs at all.
Since each worker thread processes its requests sequentially, handlers which block the current thread will cause the current worker to stop processing new requests:
fn my_handler() -> impl Responder {
std::thread::sleep(Duration::from_secs(5)); // <-- Bad practice! Will cause the current worker thread to hang!
"response"
}
For this reason, any long, non-cpu-bound operation (e.g. I/O, database operations, etc.) should be expressed as futures or asynchronous functions. Async handlers get executed concurrently by worker threads and thus don't block execution:
async fn my_handler() -> impl Responder {
tokio::time::delay_for(Duration::from_secs(5)).await; // <-- Ok. Worker thread will handle other requests here
"response"
}
The same limitation applies to extractors as well. When a handler function receives an argument
which implements FromRequest
, and that implementation blocks the current thread, the worker thread
will block when running the handler. Special attention must be given when implementing extractors
for this very reason, and they should also be implemented asynchronously where needed.
SSL
There are two features for the ssl server: rustls
and openssl
. The rustls
feature is for
rustls
integration and openssl
is for openssl
.
[dependencies]
actix-web = { version = "{{< actix-version "actix-web" >}}", features = ["openssl"] }
openssl = { version="0.10" }
{{< include-example example="server" file="ssl.rs" section="ssl" >}}
Note
: the HTTP/2.0 protocol requires tls alpn. At the moment, only
openssl
hasalpn
support. For a full example, check out examples/openssl.
To create the key.pem and cert.pem use the command. Fill in your own subject
$ openssl req -x509 -newkey rsa:4096 -keyout key.pem -out cert.pem \
-days 365 -sha256 -subj "/C=CN/ST=Fujian/L=Xiamen/O=TVlinux/OU=Org/CN=muro.lxd"
To remove the password, then copy nopass.pem to key.pem
$ openssl rsa -in key.pem -out nopass.pem
Keep-Alive
Actix can wait for requests on a keep-alive connection.
keep alive connection behavior is defined by server settings.
75
,Some(75)
,KeepAlive::Timeout(75)
- enable 75 second keep alive timer.None
orKeepAlive::Disabled
- disable keep alive.KeepAlive::Tcp(75)
- useSO_KEEPALIVE
socket option.
{{< include-example example="server" file="keep_alive.rs" section="keep-alive" >}}
If the first option above is selected, then keep alive state is calculated based on the
response's connection-type. By default HttpResponse::connection_type
is not
defined. In that case keep alive is defined by the request's http version.
keep alive is off for HTTP/1.0 and is on for HTTP/1.1 and HTTP/2.0.
Connection type can be changed with HttpResponseBuilder::connection_type()
method.
{{< include-example example="server" file="keep_alive_tp.rs" section="example" >}}
Graceful shutdown
HttpServer
supports graceful shutdown. After receiving a stop signal, workers
have a specific amount of time to finish serving requests. Any workers still alive after the
timeout are force-dropped. By default the shutdown timeout is set to 30 seconds. You
can change this parameter with the HttpServer::shutdown_timeout()
method.
You can send a stop message to the server with the server address and specify if you want
graceful shutdown or not. The start()
method returns address of the server.
HttpServer
handles several OS signals. CTRL-C is available on all OSs, other signals
are available on unix systems.
- SIGINT - Force shutdown workers
- SIGTERM - Graceful shutdown workers
- SIGQUIT - Force shutdown workers
It is possible to disable signal handling with
HttpServer::disable_signals()
method.