1. 用法: 设置超时时间
OkHttpClient httpClient = new OkHttpClient.Builder()
.retryOnConnectionFailure(true)
.connectTimeout(CONNECT_TIMEOUT, TimeUnit.SECONDS) //连接超时
.readTimeout(READ_TIMEOUT, TimeUnit.SECONDS) //读取超时
.writeTimeout(WRITE_TIMEOUT, TimeUnit.SECONDS) //写超时
.addInterceptor(new CommonHeaderInterceptor())
.addInterceptor(new CacheInterceptor())
.addInterceptor(new HttpLoggerInterceptor())
.addNetworkInterceptor(new EncryptInterceptor())
.build();
这个都知道, 一搜一大把, 但是没人讲这三种timeout有什么区别...
2. 总结
源码分析之前先上总结
- connectTimeout 最终设置给了socket (确切的说应该是rawSocket)
- readTimeout 最终设置给了rawSocket 以及 在socket基础上创建的BufferedSource
- writeTimeout 最终设置给了在socket基础上创建的BufferedSink
一言以蔽之: okhttp底层基于socket, 所以 Timeout 自然也是设置给Socket 的 connect / read / write。而socket是对于传输层的抽象, 因为我们这里讨论的是http, 所以对socket设置各种timeout 其实也就是对于TCP的timeout进行配置;
TCP协议(握手/挥手/发包/丢包重传/滑动窗口/拥塞控制等细节)以及socket属于前置知识, 若不太了解,建议先恶补一下。
以下的源码探究就是罗列记录一下自己的探究过程, 嫌啰嗦可以忽略~
3. 源码探究
我们知道 okhttp
采用了责任链的设计模式,用一条抽象的 Chain
将一堆 Interceptor
串起来,从发出request 到接收response的路径类似于node.js
中koa2
的“洋葱模型”(图1),而 okhttp
的 Interceptor
作用就相当于koa2
中的 middleware
.
“洋葱”的每一层都是一个Interceptor
,每一层都专注于自己的事情(单一职责),比如日志、mock api,弱网模拟,统一header,APP层缓存、通讯加密等,功能拆分,互不影响,从框架层面来讲也是对AOP思想的具体实践。(AOP可不仅仅是传统意义上的字节码插桩)
okhttp本身已经提供了几个Interceptor
的默认实现,比如 CacheInterceptor
就是对于http1.1缓存机制的具体实现(cache-controll等); ConnectInterceptor
专门负责创建/复用TCP连接, 里面的ConnectionPool
就是对http1.1
中 keep-alive
(TCP连接复用)和 pipline
机制(用多条TCP连接实现并发请求)的具体实现。而超时相关的设置也是从这里切入。
/** Opens a connection to the target server and proceeds to the next interceptor. */
public final class ConnectInterceptor implements Interceptor {
public final OkHttpClient client;
public ConnectInterceptor(OkHttpClient client) {
this.client = client;
}
@Override public Response intercept(Chain chain) throws IOException {
RealInterceptorChain realChain = (RealInterceptorChain) chain;
Request request = realChain.request();
StreamAllocation streamAllocation = realChain.streamAllocation();
// We need the network to satisfy this request. Possibly for validating a conditional GET.
boolean doExtensiveHealthChecks = !request.method().equals("GET");
// 入口在 newStream 方法
HttpCodec httpCodec = streamAllocation.newStream(client, chain, doExtensiveHealthChecks);
RealConnection connection = streamAllocation.connection();
return realChain.proceed(request, streamAllocation, httpCodec, connection);
}
}
上面的 StreamAllocation#newStream 方法就做了两件事,
public HttpCodec newStream(OkHttpClient client, Interceptor.Chain chain, boolean doExtensiveHealthChecks) {
// 这里的chain就是RealInterceptorChain,它里面的各种timeout值都是通过我们创建HttpClient时原封不动赋给它的,下面只是它的一些get方法;
int connectTimeout = chain.connectTimeoutMillis();
int readTimeout = chain.readTimeoutMillis();
int writeTimeout = chain.writeTimeoutMillis();
int pingIntervalMillis = client.pingIntervalMillis();
boolean connectionRetryEnabled = client.retryOnConnectionFailure();
//简化后的代码
...
// 3.1 findHealthyConnection 会调用 findConnection
RealConnection resultConnection = findHealthyConnection(connectTimeout, readTimeout,
writeTimeout, pingIntervalMillis, connectionRetryEnabled, doExtensiveHealthChecks);
// 3.2
HttpCodec resultCodec = resultConnection.newCodec(client, chain, this);
return resultCodec;
}
3.1 设置给rawSocket 上的 connectTimeout 和 readTimeout
StreamAllocation#findConnection主要做了两件事,先是从连接池中复用或者创建一个新的连接(RealConnection),然后调用 RealConnection#connect 方法完成 TCP + TLS 握手,其中TCP握手是在
RealConnection#connectSocket(connectTimeout, readTimeout, call, eventListener);
中发起的。
/**
* Does all the work necessary to build a full HTTP or HTTPS connection on a raw socket.
*/
private void connectSocket(int connectTimeout, int readTimeout, Call call,
EventListener eventListener) throws IOException {
Proxy proxy = route.proxy();
Address address = route.address();
//创建一个socket。 在未设置proxy的情况下, 会采用默认的proxySelector, 此时的proxy.type == DIRECT 即直连
rawSocket = proxy.type() == Proxy.Type.DIRECT || proxy.type() == Proxy.Type.HTTP
? address.socketFactory().createSocket() // 走这里, 实际new Socket()
: new Socket(proxy);
eventListener.connectStart(call, route.socketAddress(), proxy);
//给socket设置读取server端数据的超时;
rawSocket.setSoTimeout(readTimeout);
try {
//实际调用的是 rawSocket.connect(route.socketAddress(), connectTimeout), 建立TCP连接,同时设置连接超时
Platform.get().connectSocket(rawSocket, route.socketAddress(), connectTimeout);
} catch (ConnectException e) {
...
throw ce;
}
// The following try/catch block is a pseudo hacky way to get around a crash on Android 7.0
// More details:
// https://github.com/square/okhttp/issues/3245
// https://android-review.googlesource.com/#/c/271775/
try {
//创建source
source = Okio.buffer(Okio.source(rawSocket));
//创建sink
sink = Okio.buffer(Okio.sink(rawSocket));
} catch (NullPointerException npe) {
...
}
}
关于socket.setSoTimeout
, 以下是原文档说明的个人翻译及理解
调用此方法设置一个非0的timeout,那么调用InputStream(与此Socket相关联的) 的read()这个阻塞方法读取server端的数据时, 持续timeout之久。
如果timeout 到期,不管Socket是否有效, 都会抛出java.net.SocketTimeoutException。
这个timeout 必须在socket进入block操作之前设置 才能生效;
正常设置timeout >0, 如果设置timeout=0, 则代表 timeout无限;
关于socket.connect(address, connectTimeout);
Connects this socket to the server with a specified timeout value. A timeout of zero is interpreted as an infinite timeout. The connection will then block until established or an error occurs.
简言之就是 与server建立连接的最大时长
3.2 BufferedSource上的 readTimeout 和 BufferedSink上的writeTimeout
RealConnection#newCodec()
根据 connection 创建httpCodec(Encodes HTTP requests and decodes HTTP responses.)
public HttpCodec newCodec(OkHttpClient client, Interceptor.Chain chain, StreamAllocation streamAllocation) throws SocketException {
//前面是HTTP2相关的实现,暂略
...
//此处又给socket设置了一次readTimeout, 当然此socket已经不一定是rawSocket了
socket.setSoTimeout(chain.readTimeoutMillis());
//
source.timeout().timeout(chain.readTimeoutMillis(), MILLISECONDS);
//
sink.timeout().timeout(chain.writeTimeoutMillis(), MILLISECONDS);
return new Http1Codec(client, streamAllocation, source, sink);
}
当然还有一个地方是在connectTunnel()
用到, 但是这个前提是走http代理的时候, 这个暂且不详细探究;
3.3 下面是source和sink中的timeout 的详细解释
Source 和 Sink 是 okio 中定义的两个接口, 这两个接口都支持读写超时设置
其中source可以理解为inputstream, sink可以理解为outputstream
具体是什么鬼, 看一下source和sink的创建就是知道了
BufferedSource的创建
罗列细节之前先总结一下流程:
Socket ----> InputStream ---> Source ---> BufferedSource
还是RealConnection的connectSocket方法
//创建BufferedSource
source = Okio.buffer(Okio.source(rawSocket));
Okio.buffer(Source source)
就是new RealBufferedSource(source);
那么下面主要来看Okio.source(rawSocket)
public static Source source(Socket socket) throws IOException {
if (socket == null) throw new IllegalArgumentException("socket == null");
AsyncTimeout timeout = timeout(socket);
//此处用socket的inputstream创建了source
Source source = source(socket.getInputStream(), timeout);
return timeout.source(source);
}
//下面请看 okio 是如何将 inputstream 封装成 source 的
private static Source source(final InputStream in, final Timeout timeout) {
if (in == null) throw new IllegalArgumentException("in == null");
if (timeout == null) throw new IllegalArgumentException("timeout == null");
return new Source() {
@Override public long read(Buffer sink, long byteCount) throws IOException {
if (byteCount < 0) throw new IllegalArgumentException("byteCount < 0: " + byteCount);
if (byteCount == 0) return 0;
try {
//每次read都会检测timeout
timeout.throwIfReached();
Segment tail = sink.writableSegment(1);
int maxToCopy = (int) Math.min(byteCount, Segment.SIZE - tail.limit);
//本质还是调用了inputstream的read方法
int bytesRead = in.read(tail.data, tail.limit, maxToCopy);
if (bytesRead == -1) return -1;
tail.limit += bytesRead;
sink.size += bytesRead;
return bytesRead;
} catch (AssertionError e) {
if (isAndroidGetsocknameError(e)) throw new IOException(e);
throw e;
}
}
@Override public void close() throws IOException {
in.close();
}
@Override public Timeout timeout() {
return timeout;
}
@Override public String toString() {
return "source(" + in + ")";
}
};
}
BufferedSink的创建
跟BuffedSource很相似, 简略描述
sink = Okio.buffer(Okio.sink(rawSocket));
同样主要看Okio.sink(rawSocket)
的实现
public static Sink sink(Socket socket) throws IOException {
if (socket == null) throw new IllegalArgumentException("socket == null");
AsyncTimeout timeout = timeout(socket);
//用socket的outputstream创建sink
Sink sink = sink(socket.getOutputStream(), timeout);
return timeout.sink(sink);
}
sink静态方法的实现
private static Sink sink(final OutputStream out, final Timeout timeout) {
if (out == null) throw new IllegalArgumentException("out == null");
if (timeout == null) throw new IllegalArgumentException("timeout == null");
return new Sink() {
@Override public void write(Buffer source, long byteCount) throws IOException {
checkOffsetAndCount(source.size, 0, byteCount);
while (byteCount > 0) {
//每次write之前检测timeout
timeout.throwIfReached();
Segment head = source.head;
int toCopy = (int) Math.min(byteCount, head.limit - head.pos);
//最终调用outputstream的write方法
out.write(head.data, head.pos, toCopy);
head.pos += toCopy;
byteCount -= toCopy;
source.size -= toCopy;
if (head.pos == head.limit) {
source.head = head.pop();
SegmentPool.recycle(head);
}
}
}
@Override public void flush() throws IOException {
out.flush();
}
@Override public void close() throws IOException {
out.close();
}
@Override public Timeout timeout() {
return timeout;
}
@Override public String toString() {
return "sink(" + out + ")";
}
};
}
以上~