在默认大多数情况下，AcitveMQ 是以异步模式发送消息。例外的情况：在没有使用
事务的情况下，生产者以PERSISTENT 传送模式发送消息。在这种情况下，send 方法都
是同步的，并且一直阻塞直到ActiveMQ 发回确认消息：消息已经存储在持久性数据存
储中。这种确认机制保证消息不会丢失，但会造成生产者阻塞从而影响反应时间。
高性能的程序一般都能容忍在故障情况下丢失少量数据。如果编写这样的程序，可
以通过使用异步发送来提高吞吐量（甚至在使用PERSISTENT 传送模式的情况下）。
protected void send(ActiveMQMessageProducer producer, ActiveMQDestination destination, Message message, int deliveryMode, int priority, long timeToLive,
MemoryUsage producerWindow, int sendTimeout, AsyncCallback onComplete) throws JMSException {
checkClosed();
if (destination.isTemporary() && connection.isDeleted(destination)) {
throw new InvalidDestinationException("Cannot publish to a deleted Destination: " + destination);
}
synchronized (sendMutex) {
// tell the Broker we are about to start a new transaction
doStartTransaction();
TransactionId txid = transactionContext.getTransactionId();
long sequenceNumber = producer.getMessageSequence();
//Set the "JMS" header fields on the original message, see 1.1 spec section 3.4.11
message.setJMSDeliveryMode(deliveryMode);
long expiration = 0L;
if (!producer.getDisableMessageTimestamp()) {
long timeStamp = System.currentTimeMillis();
message.setJMSTimestamp(timeStamp);
if (timeToLive > 0) {
expiration = timeToLive + timeStamp;
}
}
message.setJMSExpiration(expiration);
message.setJMSPriority(priority);
message.setJMSRedelivered(false);
// transform to our own message format here
ActiveMQMessage msg = ActiveMQMessageTransformation.transformMessage(message, connection);
msg.setDestination(destination);
msg.setMessageId(new MessageId(producer.getProducerInfo().getProducerId(), sequenceNumber));
// Set the message id.
if (msg != message) {
message.setJMSMessageID(msg.getMessageId().toString());
// Make sure the JMS destination is set on the foreign messages too.
message.setJMSDestination(destination);
}
//clear the brokerPath in case we are re-sending this message
msg.setBrokerPath(null);
msg.setTransactionId(txid);
if (connection.isCopyMessageOnSend()) {
msg = (ActiveMQMessage)msg.copy();
}
msg.setConnection(connection);
msg.onSend();
msg.setProducerId(msg.getMessageId().getProducerId());
if (LOG.isTraceEnabled()) {
LOG.trace(getSessionId() + " sending message: " + msg);
}
// 请看下面这一行就是判断是同步发送还是异步发送的。
if (onComplete==null && sendTimeout <= 0 && !msg.isResponseRequired() && !connection.isAlwaysSyncSend() && (!msg.isPersistent() || connection.isUseAsyncSend() || txid != null)) {
this.connection.asyncSendPacket(msg);
if (producerWindow != null) {
// Since we defer lots of the marshaling till we hit the
// wire, this might not
// provide and accurate size. We may change over to doing
// more aggressive marshaling,
// to get more accurate sizes.. this is more important once
// users start using producer window
// flow control.
int size = msg.getSize();
producerWindow.increaseUsage(size);
}
} else {
if (sendTimeout > 0 && onComplete==null) {
this.connection.syncSendPacket(msg,sendTimeout);
}else {
this.connection.syncSendPacket(msg, onComplete);
}
}
}
}
接下来看看同步发送方法。
public Object request(Object command) throws IOException {
FutureResponse response = asyncRequest(command, null); // 调用发送方法
return response.getResult(); // 从future方法阻塞等待返回
}
看下asyncRequest方法。构建了一个FutureResponse对象。并且将其存到了
requestMap.put(new Integer(command.getCommandId()), future);这个map当中。
等broker有返回的时候，从requestMap里面取出来这个future.
public FutureResponse asyncRequest(Object o, ResponseCallback responseCallback) throws IOException {
Command command = (Command) o;
command.setCommandId(sequenceGenerator.getNextSequenceId());
command.setResponseRequired(true);
FutureResponse future = new FutureResponse(responseCallback, this);
IOException priorError = null;
synchronized (requestMap) {
priorError = this.error;
if (priorError == null) {
requestMap.put(new Integer(command.getCommandId()), future);
}
}
if (priorError != null) {
future.set(new ExceptionResponse(priorError));
throw priorError;
}
next.oneway(command);
return future;
}看下FutureResponse里面的这个回调方法。
public void set(Response result) {
if (responseSlot.offer(result)) {
if (responseCallback != null) {
responseCallback.onCompletion(this);
}
}
}其中responseSlot是个阻塞队列 private final ArrayBlockingQueue responseSlot = new ArrayBlockingQueue(1);
。复习下阻塞队列的知识。
使用BlockingQueue的关键技术点如下:
1.BlockingQueue定义的常用方法如下:
1)add(anObject):把anObject加到BlockingQueue里,即如果BlockingQueue可以容纳,则返回true,否则报异常
2)offer(anObject):表示如果可能的话,将anObject加到BlockingQueue里,即如果BlockingQueue可以容纳,则返回true,否则返回false.
3)put(anObject):把anObject加到BlockingQueue里,如果BlockQueue没有空间,则调用此方法的线程被阻断直到BlockingQueue里面有空间再继续.
4)poll(time):取走BlockingQueue里排在首位的对象,若不能立即取出,则可以等time参数规定的时间,取不到时返回null
5)take():取走BlockingQueue里排在首位的对象,若BlockingQueue为空,阻断进入等待状态直到Blocking有新的对象被加入为止
2.BlockingQueue有四个具体的实现类,根据不同需求,选择不同的实现类
1)ArrayBlockingQueue:规定大小的BlockingQueue,其构造函数必须带一个int参数来指明其大小.其所含的对象是以FIFO(先入先出)顺序排序的.
2)LinkedBlockingQueue:大小不定的BlockingQueue,若其构造函数带一个规定大小的参数,生成的BlockingQueue有大小限制,若不带大小参数,所生成的BlockingQueue的大小由Integer.MAX_VALUE来决定.其所含的对象是以FIFO(先入先出)顺序排序的
3)PriorityBlockingQueue:类似于LinkedBlockQueue,但其所含对象的排序不是FIFO,而是依据对象的自然排序顺序或者是构造函数的Comparator决定的顺序.
4)SynchronousQueue:特殊的BlockingQueue,对其的操作必须是放和取交替完成的.
3.LinkedBlockingQueue和ArrayBlockingQueue比较起来,它们背后所用的数据结构不一样,导致LinkedBlockingQueue的数据吞吐量要大于ArrayBlockingQueue,但在线程数量很大时其性能的可预见性低于ArrayBlockingQueue.