OP_WRITE事件是在Socket发送缓冲区中的可用字节数大于或等于其低水位标记SO_SNDLOWAT时发生。正常情况下,都是可写的,因此一般不注册写事件。所以一般代码如下:
while (bb.hasRemaining()) {
int len = socketChannel.write(bb);
if (len < 0) {
throw new EOFException();
}
}
这样在大部分情况都没问题,但是高并发,并且在网络环境很差的情况下,发送缓冲区可能会满,导致无限循环,这样最终会导致CPU利用率100%。下面就看看一些基于NIO的框架,是如何处理这个问题的。
private void handleWrites(SelectionKey sk, MemcachedNode qa)
throws IOException {
// 填充写缓冲区
qa.fillWriteBuffer(shouldOptimize);
boolean canWriteMore = qa.getBytesRemainingToWrite() > 0;
while (canWriteMore) {
int wrote = qa.writeSome();
qa.fillWriteBuffer(shouldOptimize);
// 如果wrote等于零,表示没有写出数据,那么不再尝试写,等待下次线程外层循环注册write事件
canWriteMore = wrote > 0 && qa.getBytesRemainingToWrite() > 0;
}
public final int writeSome() throws IOException {
int wrote = channel.write(wbuf);
// 写入多少个字节,toWrite就减去对应的数量
toWrite -= wrote;
return wrote;
}
public final int getSelectionOps() {
int rv = 0;
if (getChannel().isConnected()) {
if (hasReadOp()) {
rv |= SelectionKey.OP_READ;
}
// 如果toWrite大于0,说明由于某种异常原因上次写入还未完成;hasWriteOp()用于判断写队列是否还有元素。这两种情况下,需要注册写事件。本文讨论的是toWrite>0的情况。
if (toWrite > 0 || hasWriteOp()) {
rv |= SelectionKey.OP_WRITE;
}
} else {
rv = SelectionKey.OP_CONNECT;
}
return rv;
}
说明:Spymemcached是单线程的,因此就是绝对不能阻塞,所以当发现不可写的时候,不能阻塞住线程,而是立即返回,等待下次主线程循环来注册事件。
protected void write0(AbstractNioChannel> channel) {
boolean open = true;
boolean addOpWrite = false;
boolean removeOpWrite = false;
boolean iothread = isIoThread(channel);
long writtenBytes = 0;
final SocketSendBufferPool sendBufferPool = this.sendBufferPool;
final WritableByteChannel ch = channel.channel;
final Queue writeBuffer = channel.writeBufferQueue;
final int writeSpinCount = channel.getConfig().getWriteSpinCount();
List causes = null;
synchronized (channel.writeLock) {
channel.inWriteNowLoop = true;
for (;;) {
MessageEvent evt = channel.currentWriteEvent;
SendBuffer buf = null;
ChannelFuture future = null;
try {
if (evt == null) {
if ((channel.currentWriteEvent = evt = writeBuffer.poll()) == null) {
// 如果无数据可写,则需要删除可写事件的注册
removeOpWrite = true;
channel.writeSuspended = false;
break;
}
future = evt.getFuture();
channel.currentWriteBuffer = buf = sendBufferPool.acquire(evt.getMessage());
} else {
future = evt.getFuture();
buf = channel.currentWriteBuffer;
}
long localWrittenBytes = 0;
// 通过writeSpinCount来控制尝试写的次数,如果最终还是无法写入,就注册写事件
for (int i = writeSpinCount; i > 0; i --) {
// 写数据
localWrittenBytes = buf.transferTo(ch);
// 如果写入数据不等于零,表明写入成功,跳出循环
if (localWrittenBytes != 0) {
writtenBytes += localWrittenBytes;
break;
}
// 如果buf的数据都写完了,则跳出循环
if (buf.finished()) {
break;
}
}
if (buf.finished()) {
// Successful write - proceed to the next message.
buf.release();
channel.currentWriteEvent = null;
channel.currentWriteBuffer = null;
// Mark the event object for garbage collection.
//noinspection UnusedAssignment
evt = null;
buf = null;
future.setSuccess();
} else {
// Not written fully - perhaps the kernel buffer is full.
addOpWrite = true;
channel.writeSuspended = true;
if (writtenBytes > 0) {
// Notify progress listeners if necessary.
future.setProgress(
localWrittenBytes,
buf.writtenBytes(), buf.totalBytes());
}
break;
}
}
}
channel.inWriteNowLoop = false;
if (open) {
if (addOpWrite) {
// 注册写事件
setOpWrite(channel);
} else if (removeOpWrite) {
// 删除写事件
clearOpWrite(channel);
}
}
}
}
说明:Netty是多线程的,因此其可以通过阻塞线程做一定的等待,等待通道可写。Netty等待是通过spinCount等待指定的循环次数。
public static long flushChannel(SocketChannel socketChannel, ByteBuffer bb, long writeTimeout)
throws IOException {
SelectionKey key = null;
Selector writeSelector = null;
int attempts = 0;
int bytesProduced = 0;
try {
while (bb.hasRemaining()) {
int len = socketChannel.write(bb);
// 类似Netty的spinCount
attempts++;
if (len < 0) {
throw new EOFException();
}
bytesProduced += len;
if (len == 0) {
if (writeSelector == null) {
// 获取一个新的selector
writeSelector = SelectorFactory.getSelector();
if (writeSelector == null) {
// Continue using the main one
continue;
}
}
// 在新selector上注册写事件,而不是在主selector上注册
key = socketChannel.register(writeSelector, key.OP_WRITE);
// 利用writeSelector.select()来阻塞当前线程,等待可写事件发生,总共等待可写事件的时长是3*writeTimeout
if (writeSelector.select(writeTimeout) == 0) {
if (attempts > 2)
throw new IOException("Client disconnected");
} else {
attempts--;
}
} else {
attempts = 0;
}
}
}
return bytesProduced;
}
说明:Grizzly是多线程的,因此其可以做合适的阻塞等待。其没有再主selector上注册写事件,而是在重新构造的selector上注册写事件,并且通过select()来阻塞一定的时间来等待可写。
为什么要这么做呢?Grizzly的作者对此的回应如下:
1. 使用临时的Selector的目的是减少线程间的切换。当前的Selector一般用来处理OP_ACCEPT,和OP_READ的操作。使用临时的Selector可减轻主Selector的负担;而在注册的时候则需要进行线程切换,会引起不必要的系统调用。这种方式避免了线程之间的频繁切换,有利于系统的性能提高。
2. 虽然writeSelector.select(writeTimeout)做了阻塞操作,但是这种情况只是少数极端的环境下才会发生。> 大多数的客户端是不会频繁出现这种现象的,因此在同一时刻被阻塞的线程不会很多。
3. 利用这个阻塞操作来判断异常中断的客户连接。
4. 经过压力实验证明这种实现的性能是非常好的。