OkHttp请求耗时统计和实践

2020-09-23 本文已影响0人杨充211

目录介绍

01.先提问一个问题
02.EventListener回调原理
03.请求开始结束监听
04.dns解析开始结束监听
05.连接开始结束监听
06.TLS连接开始结束监听
07.连接绑定和释放监听
08.request请求监听
09.response响应监听
10.如何监听统计耗时
11.应用实践之案例

01.先提问一个问题

OkHttp如何进行各个请求环节的耗时统计呢？
- OkHttp 版本提供了EventListener接口，可以让调用者接收一系列网络请求过程中的事件，例如DNS解析、TSL/SSL连接、Response接收等。
- 通过继承此接口，调用者可以监视整个应用中网络请求次数、流量大小、耗时(比如dns解析时间，请求时间，响应时间等等)情况。

02.EventListener回调原理

先来看一下

public abstract class EventListener {
   // 按照请求顺序回调
    public void callStart(Call call) {}
    // 域名解析
    public void dnsStart(Call call, String domainName) {}
    public void dnsEnd(Call call, String domainName, List<InetAddress> inetAddressList) {}
    // 释放当前Transmitter的RealConnection
    public void connectionReleased(Call call, Connection connection) {}
    public void connectionAcquired(call, result){};
    // 开始连接
    public void connectStart(call, route.socketAddress(), proxy){}
    // 请求
    public void requestHeadersStart(@NotNull Call call){}
    public void requestHeadersEnd(@NotNull Call call, @NotNull Request request) {}
    // 响应
    public void requestBodyStart(@NotNull Call call) {}
    public void requestBodyEnd(@NotNull Call call, long byteCount) {}
    // 结束
    public void callEnd(Call call) {}
    // 失败
    public void callFailed(Call call, IOException ioe) {}
}

03.请求开始结束监听

callStart(Call call) 请求开始

当一个Call（代表一个请求）被同步执行或被添加异步队列中时，即会调用这个回调方法。
需要说明这个方法是在dispatcher.executed/enqueue前执行的。
由于线程或事件流的限制，这里的请求开始并不是真正的去执行的这个请求。如果发生重定向和多域名重试时，这个方法也仅被调用一次。

final class RealCall implements Call {
    @Override 
    public Response execute() throws IOException {
        eventListener.callStart(this);
        client.dispatcher().executed(this);
        Response result = getResponseWithInterceptorChain();
        if (result == null) throw new IOException("Canceled");
        return result;    
    }

    @Override 
    public void enqueue(Callback responseCallback) {
        eventListener.callStart(this);
        client.dispatcher().enqueue(new AsyncCall(responseCallback));
    }
}

callFailed/callEnd 请求异常和请求结束

每一个callStart都对应着一个callFailed或callEnd。
callFailed在两种情况下被调用，第一种是在请求执行的过程中发生异常时。第二种是在请求结束后，关闭输入流时产生异常时。

final class RealCall implements Call {
    @Override 
    public Response execute() throws IOException {
        try {
          client.dispatcher().executed(this);
          Response result = getResponseWithInterceptorChain();
          if (result == null) throw new IOException("Canceled");
          return result;
        } catch (IOException e) {
          eventListener.callFailed(this, e);
          throw e;
        }
    }
    final class AsyncCall extends NamedRunnable {
        @Override 
        protected void execute() {
            try {
                Response response = getResponseWithInterceptorChain();
            } catch (IOException e) {
                eventListener.callFailed(RealCall.this, e);

            }
        }
    }
}

//第二种
public final class StreamAllocation {
    public void streamFinished(boolean noNewStreams, HttpCodec codec, long bytesRead, IOException e) {
        ...
        if (e != null) {
          eventListener.callFailed(call, e);
        } else if (callEnd) {
          eventListener.callEnd(call);
        }
        ...
    }
}

callEnd也有两种调用场景。第一种也是在关闭流时。第二种是在释放连接时。

public final class StreamAllocation {

    public void streamFinished(boolean noNewStreams, HttpCodec codec, long bytesRead, IOException e) {
        ...
        if (e != null) {
          eventListener.callFailed(call, e);
        } else if (callEnd) {
          eventListener.callEnd(call);
        }
        ...
    }

    public void release() {
        ...
        if (releasedConnection != null) {
          eventListener.connectionReleased(call, releasedConnection);
          eventListener.callEnd(call);
        }
    }
}

为什么会将关闭流和关闭连接区分开？
- 在http2版本中，一个连接上允许打开多个流，OkHttp使用StreamAllocation来作为流和连接的桥梁。当一个流被关闭时，要检查这条连接上还有没有其他流，如果没有其他流了，则可以将连接关闭了。
- streamFinished和release作用是一样的，都是关闭当前流，并检查是否需要关闭连接。不同的是，当调用者手动取消请求时，调用的是release方法，并由调用者负责关闭请求输出流和响应输入流。

04.dns解析开始结束监听

dnsStart开始

其中的lookup(String hostname)方法代表了域名解析的过程，dnsStart/dnsEnd就是在lookup前后被调用的
DNS解析是请求DNS（Domain Name System）服务器，将域名解析成ip的过程。域名解析工作是由JDK中的InetAddress类完成的。

  /** Prepares the socket addresses to attempt for the current proxy or host. */
  private void resetNextInetSocketAddress(Proxy proxy) throws IOException {
    if (proxy.type() == Proxy.Type.SOCKS) {
      inetSocketAddresses.add(InetSocketAddress.createUnresolved(socketHost, socketPort));
    } else {
      eventListener.dnsStart(call, socketHost);

      // Try each address for best behavior in mixed IPv4/IPv6 environments.
      List<InetAddress> addresses = address.dns().lookup(socketHost);
      if (addresses.isEmpty()) {
        throw new UnknownHostException(address.dns() + " returned no addresses for " + socketHost);
      }

      eventListener.dnsEnd(call, socketHost, addresses);
    }
  }

那么RouteSelector这个类是在哪里调用

public final class StreamAllocation {

  public StreamAllocation(ConnectionPool connectionPool, Address address, Call call,
      EventListener eventListener, Object callStackTrace) {
    this.routeSelector = new RouteSelector(address, routeDatabase(), call, eventListener);
  }
}

05.连接开始结束监听

connectStart连接开始

OkHttp是使用Socket接口建立Tcp连接的，所以这里的连接就是指Socket建立一个连接的过程。
当连接被重用时，connectStart/connectEnd不会被调用。当请求被重定向到新的域名后，connectStart/connectEnd会被调用多次。

  private void connectSocket(int connectTimeout, int readTimeout, Call call,
      EventListener eventListener) throws IOException {
    Proxy proxy = route.proxy();
    Address address = route.address();

    rawSocket = proxy.type() == Proxy.Type.DIRECT || proxy.type() == Proxy.Type.HTTP
        ? address.socketFactory().createSocket()
        : new Socket(proxy);

    eventListener.connectStart(call, route.socketAddress(), proxy);
  }

connectEnd连接结束

因为创建的连接有两种类型（服务端直连和隧道代理），所以callEnd有两处调用位置。为了在基于代理的连接上使用SSL，需要单独发送CONECT请求。
在连接过程中，无论是Socket连接失败，还是TSL/SSL握手失败，都会回调connectEnd。

  public void connect(int connectTimeout, int readTimeout, int writeTimeout,
    while (true) {
      try {
        establishProtocol(connectionSpecSelector, pingIntervalMillis, call, eventListener);
        eventListener.connectEnd(call, route.socketAddress(), route.proxy(), protocol);
        break;
      } catch (IOException e) {
        eventListener.connectFailed(call, route.socketAddress(), route.proxy(), null, e);
      }
  }

  private void connectTunnel(int connectTimeout, int readTimeout, int writeTimeout, Call call,
      EventListener eventListener) throws IOException {
    Request tunnelRequest = createTunnelRequest();
    HttpUrl url = tunnelRequest.url();
    for (int i = 0; i < MAX_TUNNEL_ATTEMPTS; i++) {
      connectSocket(connectTimeout, readTimeout, call, eventListener);
      eventListener.connectEnd(call, route.socketAddress(), route.proxy(), null);
    }
  }

06.TLS连接开始结束监听

开始连接，代码如下所示

在上面看到，在Socket建立连接后，会执行一个establishProtocol方法，这个方法的作用就是TSL/SSL握手。
当存在重定向或连接重试的情况下，secureConnectStart/secureConnectEnd会被调用多次。

  private void establishProtocol(ConnectionSpecSelector connectionSpecSelector,
      int pingIntervalMillis, Call call, EventListener eventListener) throws IOException {
    if (route.address().sslSocketFactory() == null) {
      protocol = Protocol.HTTP_1_1;
      socket = rawSocket;
      return;
    }

    eventListener.secureConnectStart(call);
    connectTls(connectionSpecSelector);
    eventListener.secureConnectEnd(call, handshake);
  }

结合连接监听可知
- 如果我们使用了HTTPS安全连接，在TCP连接成功后需要进行TLS安全协议通信，等TLS通讯结束后才能算是整个连接过程的结束，也就是说connectEnd在secureConnectEnd之后调用。
所以顺序是这样的
- connectStart ---> secureConnectStart ---> secureConnectEnd ---> ConnectEnd

07.连接绑定和释放监听

因为OkHttp是基于连接复用的，当一次请求结束后并不会马上关闭当前连接，而是放到连接池中。

当有相同域名的请求时，会从连接池中取出对应的连接使用，减少了连接的频繁创建和销毁。
当根据一个请求从连接池取连接时，并打开输入输出流就是acquired，用完释放流就是released。
如果直接复用StreamAllocation中的连接，则不会调用connectionAcquired/connectReleased。

  private RealConnection findConnection(int connectTimeout, int readTimeout, int writeTimeout,
      int pingIntervalMillis, boolean connectionRetryEnabled) throws IOException {
    synchronized (connectionPool) {
      if (result == null) {
        // 第一次查缓存 Attempt to get a connection from the pool.
        // Attempt to get a connection from the pool.
        Internal.instance.get(connectionPool, address, this, null);
      }
    }

    if (releasedConnection != null) {
      eventListener.connectionReleased(call, releasedConnection);
    }
    if (foundPooledConnection) {
      eventListener.connectionAcquired(call, result);
    }

    synchronized (connectionPool) {
      if (canceled) throw new IOException("Canceled");

      if (newRouteSelection) {
        //第二次查缓存
        List<Route> routes = routeSelection.getAll();
        for (int i = 0, size = routes.size(); i < size; i++) {
          Route route = routes.get(i);
          Internal.instance.get(connectionPool, address, this, route);
          if (connection != null) {
            foundPooledConnection = true;
            result = connection;
            this.route = route;
            break;
          }
        }
      }

      if (!foundPooledConnection) {
        //如果缓存没有，则新建连接
        route = selectedRoute;
        refusedStreamCount = 0;
        result = new RealConnection(connectionPool, selectedRoute);
        acquire(result, false);
      }
    }

    // If we found a pooled connection on the 2nd time around, we're done.
    if (foundPooledConnection) {
      eventListener.connectionAcquired(call, result);
      return result;
    }

    // Do TCP + TLS handshakes. This is a blocking operation.
    result.connect(connectTimeout, readTimeout, writeTimeout, pingIntervalMillis,
        connectionRetryEnabled, call, eventListener);
    routeDatabase().connected(result.route());

    eventListener.connectionAcquired(call, result);
    return result;
  }

connectionAcquired是在连接成功后被调用的。
- 但是在连接复用的情况下没有连接步骤，connectAcquired会在获取缓存连接后被调用。由于StreamAllocation是连接“Stream”和“Connection”的桥梁，所以在StreamAllocation中会持有一个RealConnection引用。StreamAllocation在查找可用连接的顺序为：StreamAllocation.RealConnection -> ConnectionPool -> ConnectionPool -> new RealConnection

08.request请求监听

在OkHttp中，HttpCodec负责对请求和响应按照Http协议进行编解码，包含发送请求头、发送请求体、读取响应头、读取响应体。

requestHeaders开始和结束，这个直接看CallServerInterceptor拦截器代码即可。

public final class CallServerInterceptor implements Interceptor {

  @Override public Response intercept(Chain chain) throws IOException {
    RealInterceptorChain realChain = (RealInterceptorChain) chain;
    HttpCodec httpCodec = realChain.httpStream();
    StreamAllocation streamAllocation = realChain.streamAllocation();
    RealConnection connection = (RealConnection) realChain.connection();
    Request request = realChain.request();

    long sentRequestMillis = System.currentTimeMillis();

    realChain.eventListener().requestHeadersStart(realChain.call());
    httpCodec.writeRequestHeaders(request);
    realChain.eventListener().requestHeadersEnd(realChain.call(), request);
    
    if (HttpMethod.permitsRequestBody(request.method()) && request.body() != null) {
      if (responseBuilder == null) {
        // Write the request body if the "Expect: 100-continue" expectation was met.
        realChain.eventListener().requestBodyStart(realChain.call());
        long contentLength = request.body().contentLength();
        CountingSink requestBodyOut =
            new CountingSink(httpCodec.createRequestBody(request, contentLength));
        BufferedSink bufferedRequestBody = Okio.buffer(requestBodyOut);

        request.body().writeTo(bufferedRequestBody);
        bufferedRequestBody.close();
        realChain.eventListener().requestBodyEnd(realChain.call(), requestBodyOut.successfulCount);
      } 
    }
    return response;
  }
}

09.response响应监听

responseHeadersStart和responseHeadersEnd代码如下所示

public final class CallServerInterceptor implements Interceptor {

  @Override public Response intercept(Chain chain) throws IOException {

    Response.Builder responseBuilder = null;
    if (HttpMethod.permitsRequestBody(request.method()) && request.body() != null) {
      if ("100-continue".equalsIgnoreCase(request.header("Expect"))) {
        httpCodec.flushRequest();
        realChain.eventListener().responseHeadersStart(realChain.call());
        responseBuilder = httpCodec.readResponseHeaders(true);
      }
    }

    httpCodec.finishRequest();

    if (responseBuilder == null) {
      realChain.eventListener().responseHeadersStart(realChain.call());
      responseBuilder = httpCodec.readResponseHeaders(false);
    }

    int code = response.code();
    if (code == 100) {
      // server sent a 100-continue even though we did not request one.
      // try again to read the actual response
      responseBuilder = httpCodec.readResponseHeaders(false);

      response = responseBuilder
              .request(request)
              .handshake(streamAllocation.connection().handshake())
              .sentRequestAtMillis(sentRequestMillis)
              .receivedResponseAtMillis(System.currentTimeMillis())
              .build();

      code = response.code();
    }

    realChain.eventListener() .responseHeadersEnd(realChain.call(), response);
    return response;
  }
}

responseBodyStart监听

响应体的读取有些复杂，要根据不同类型的Content-Type决定如何读取响应体，例如固定长度的、基于块(chunk)数据的、未知长度的。具体看openResponseBody方法里面的代码。
同时Http1与Http2也有不同的解析方式。下面以Http1为例。

public final class Http1Codec implements HttpCodec {

  @Override public ResponseBody openResponseBody(Response response) throws IOException {
    streamAllocation.eventListener.responseBodyStart(streamAllocation.call);
    String contentType = response.header("Content-Type");

    if (!HttpHeaders.hasBody(response)) {
      Source source = newFixedLengthSource(0);
      return new RealResponseBody(contentType, 0, Okio.buffer(source));
    }

    if ("chunked".equalsIgnoreCase(response.header("Transfer-Encoding"))) {
      Source source = newChunkedSource(response.request().url());
      return new RealResponseBody(contentType, -1L, Okio.buffer(source));
    }

    long contentLength = HttpHeaders.contentLength(response);
    if (contentLength != -1) {
      Source source = newFixedLengthSource(contentLength);
      return new RealResponseBody(contentType, contentLength, Okio.buffer(source));
    }

    return new RealResponseBody(contentType, -1L, Okio.buffer(newUnknownLengthSource()));
  }
}

responseBodyEnd监听

由下面代码可知，当响应结束后，会调用连接callEnd回调（如果异常则会调用callFailed回调）

public final class StreamAllocation {
  public void streamFinished(boolean noNewStreams, HttpCodec codec, long bytesRead, IOException e) {
    eventListener.responseBodyEnd(call, bytesRead);
    if (releasedConnection != null) {
      eventListener.connectionReleased(call, releasedConnection);
    }
    if (e != null) {
      eventListener.callFailed(call, e);
    } else if (callEnd) {
      eventListener.callEnd(call);
    }
  }
}

10.如何监听统计耗时

如何消耗记录时间
- 在OkHttp库中有一个EventListener类。该类是网络事件的侦听器。扩展这个类以监视应用程序的HTTP调用的数量、大小和持续时间。
- 所有启动/连接/获取事件最终将接收到匹配的结束/释放事件，要么成功(非空参数)，要么失败(非空可抛出)。
- 比如，可以在开始链接记录时间；dns开始，结束等方法解析记录时间，可以计算dns的解析时间。
- 比如，可以在开始请求记录时间，记录connectStart，connectEnd等方法时间，则可以计算出connect连接时间。

代码如下所示

Eventlistener只适用于没有并发的情况，如果有多个请求并发执行我们需要使用Eventlistener. Factory来给每个请求创建一个Eventlistener。
这个mRequestId是唯一值，可以选择使用AtomicInteger自增+1的方式设置id，这个使用了cas保证多线程条件下的原子性特性。

/**
 * <pre>
 *     @author yangchong
 *     email  : yangchong211@163.com
 *     time  : 2019/07/22
 *     desc  : EventListener子类
 *     revise:
 * </pre>
 */
public class NetworkListener extends EventListener {

    private static final String TAG = "NetworkEventListener";
    private static AtomicInteger mNextRequestId = new AtomicInteger(0);
    private String mRequestId ;

    public static Factory get(){
        Factory factory = new Factory() {
            @NotNull
            @Override
            public EventListener create(@NotNull Call call) {
                return new NetworkListener();
            }
        };
        return factory;
    }

    @Override
    public void callStart(@NotNull Call call) {
        super.callStart(call);
        //mRequestId = mNextRequestId.getAndIncrement() + "";
        //getAndAdd，在多线程下使用cas保证原子性
        mRequestId = String.valueOf(mNextRequestId.getAndIncrement());
        ToolLogUtils.i(TAG+"-------callStart---requestId-----"+mRequestId);
        saveEvent(NetworkTraceBean.CALL_START);
        saveUrl(call.request().url().toString());
    }

    @Override
    public void dnsStart(@NotNull Call call, @NotNull String domainName) {
        super.dnsStart(call, domainName);
        ToolLogUtils.d(TAG, "dnsStart");
        saveEvent(NetworkTraceBean.DNS_START);
    }

    @Override
    public void dnsEnd(@NotNull Call call, @NotNull String domainName, @NotNull List<InetAddress> inetAddressList) {
        super.dnsEnd(call, domainName, inetAddressList);
        ToolLogUtils.d(TAG, "dnsEnd");
        saveEvent(NetworkTraceBean.DNS_END);
    }

    @Override
    public void connectStart(@NotNull Call call, @NotNull InetSocketAddress inetSocketAddress, @NotNull Proxy proxy) {
        super.connectStart(call, inetSocketAddress, proxy);
        ToolLogUtils.d(TAG, "connectStart");
        saveEvent(NetworkTraceBean.CONNECT_START);
    }

    @Override
    public void secureConnectStart(@NotNull Call call) {
        super.secureConnectStart(call);
        ToolLogUtils.d(TAG, "secureConnectStart");
        saveEvent(NetworkTraceBean.SECURE_CONNECT_START);
    }

    @Override
    public void secureConnectEnd(@NotNull Call call, @Nullable Handshake handshake) {
        super.secureConnectEnd(call, handshake);
        ToolLogUtils.d(TAG, "secureConnectEnd");
        saveEvent(NetworkTraceBean.SECURE_CONNECT_END);
    }

    @Override
    public void connectEnd(@NotNull Call call, @NotNull InetSocketAddress inetSocketAddress,
                           @NotNull Proxy proxy, @Nullable Protocol protocol) {
        super.connectEnd(call, inetSocketAddress, proxy, protocol);
        ToolLogUtils.d(TAG, "connectEnd");
        saveEvent(NetworkTraceBean.CONNECT_END);
    }

    @Override
    public void connectFailed(@NotNull Call call, @NotNull InetSocketAddress inetSocketAddress, @NotNull Proxy proxy, @Nullable Protocol protocol, @NotNull IOException ioe) {
        super.connectFailed(call, inetSocketAddress, proxy, protocol, ioe);
        ToolLogUtils.d(TAG, "connectFailed");
    }

    @Override
    public void requestHeadersStart(@NotNull Call call) {
        super.requestHeadersStart(call);
        ToolLogUtils.d(TAG, "requestHeadersStart");
        saveEvent(NetworkTraceBean.REQUEST_HEADERS_START);
    }

    @Override
    public void requestHeadersEnd(@NotNull Call call, @NotNull Request request) {
        super.requestHeadersEnd(call, request);
        ToolLogUtils.d(TAG, "requestHeadersEnd");
        saveEvent(NetworkTraceBean.REQUEST_HEADERS_END);
    }

    @Override
    public void requestBodyStart(@NotNull Call call) {
        super.requestBodyStart(call);
        ToolLogUtils.d(TAG, "requestBodyStart");
        saveEvent(NetworkTraceBean.REQUEST_BODY_START);
    }

    @Override
    public void requestBodyEnd(@NotNull Call call, long byteCount) {
        super.requestBodyEnd(call, byteCount);
        ToolLogUtils.d(TAG, "requestBodyEnd");
        saveEvent(NetworkTraceBean.REQUEST_BODY_END);
    }

    @Override
    public void responseHeadersStart(@NotNull Call call) {
        super.responseHeadersStart(call);
        ToolLogUtils.d(TAG, "responseHeadersStart");
        saveEvent(NetworkTraceBean.RESPONSE_HEADERS_START);
    }

    @Override
    public void responseHeadersEnd(@NotNull Call call, @NotNull Response response) {
        super.responseHeadersEnd(call, response);
        ToolLogUtils.d(TAG, "responseHeadersEnd");
        saveEvent(NetworkTraceBean.RESPONSE_HEADERS_END);
    }

    @Override
    public void responseBodyStart(@NotNull Call call) {
        super.responseBodyStart(call);
        ToolLogUtils.d(TAG, "responseBodyStart");
        saveEvent(NetworkTraceBean.RESPONSE_BODY_START);
    }

    @Override
    public void responseBodyEnd(@NotNull Call call, long byteCount) {
        super.responseBodyEnd(call, byteCount);
        ToolLogUtils.d(TAG, "responseBodyEnd");
        saveEvent(NetworkTraceBean.RESPONSE_BODY_END);
    }

    @Override
    public void callEnd(@NotNull Call call) {
        super.callEnd(call);
        ToolLogUtils.d(TAG, "callEnd");
        saveEvent(NetworkTraceBean.CALL_END);
        generateTraceData();
        NetWorkUtils.timeoutChecker(mRequestId);
    }

    @Override
    public void callFailed(@NotNull Call call, @NotNull IOException ioe) {
        super.callFailed(call, ioe);
        ToolLogUtils.d(TAG, "callFailed");
    }

    private void generateTraceData(){
        NetworkTraceBean traceModel = IDataPoolHandleImpl.getInstance().getNetworkTraceModel(mRequestId);
        Map<String, Long> eventsTimeMap = traceModel.getNetworkEventsMap();
        Map<String, Long> traceList = traceModel.getTraceItemList();
        traceList.put(NetworkTraceBean.TRACE_NAME_TOTAL,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.CALL_START, NetworkTraceBean.CALL_END));
        traceList.put(NetworkTraceBean.TRACE_NAME_DNS,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.DNS_START, NetworkTraceBean.DNS_END));
        traceList.put(NetworkTraceBean.TRACE_NAME_SECURE_CONNECT,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.SECURE_CONNECT_START, NetworkTraceBean.SECURE_CONNECT_END));
        traceList.put(NetworkTraceBean.TRACE_NAME_CONNECT,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.CONNECT_START, NetworkTraceBean.CONNECT_END));
        traceList.put(NetworkTraceBean.TRACE_NAME_REQUEST_HEADERS,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.REQUEST_HEADERS_START, NetworkTraceBean.REQUEST_HEADERS_END));
        traceList.put(NetworkTraceBean.TRACE_NAME_REQUEST_BODY,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.REQUEST_BODY_START, NetworkTraceBean.REQUEST_BODY_END));
        traceList.put(NetworkTraceBean.TRACE_NAME_RESPONSE_HEADERS,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.RESPONSE_HEADERS_START, NetworkTraceBean.RESPONSE_HEADERS_END));
        traceList.put(NetworkTraceBean.TRACE_NAME_RESPONSE_BODY,NetWorkUtils.getEventCostTime(eventsTimeMap,NetworkTraceBean.RESPONSE_BODY_START, NetworkTraceBean.RESPONSE_BODY_END));
    }

    private void saveEvent(String eventName){
        NetworkTraceBean networkTraceModel = IDataPoolHandleImpl.getInstance().getNetworkTraceModel(mRequestId);
        Map<String, Long> networkEventsMap = networkTraceModel.getNetworkEventsMap();
        networkEventsMap.put(eventName, SystemClock.elapsedRealtime());
    }

    private void saveUrl(String url){
        NetworkTraceBean networkTraceModel = IDataPoolHandleImpl.getInstance().getNetworkTraceModel(mRequestId);
        networkTraceModel.setUrl(url);
    }

}

关于执行顺序，打印结果如下所示

2020-09-22 20:50:15.351 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: dnsStart
2020-09-22 20:50:15.373 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: dnsEnd
2020-09-22 20:50:15.374 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: connectStart
2020-09-22 20:50:15.404 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: secureConnectStart
2020-09-22 20:50:15.490 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: secureConnectEnd
2020-09-22 20:50:15.490 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: connectEnd
2020-09-22 20:50:15.492 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: requestHeadersStart
2020-09-22 20:50:15.492 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: requestHeadersEnd
2020-09-22 20:50:15.528 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: responseHeadersStart
2020-09-22 20:50:15.528 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: responseHeadersEnd
2020-09-22 20:50:15.532 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: responseBodyStart
2020-09-22 20:50:15.534 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: responseBodyEnd
2020-09-22 20:50:15.547 28144-28277/cn.com.zwwl.bayuwen D/NetworkEventListener: callEnd