您好,登录后才能下订单哦!
本篇内容介绍了“Netty NioEventLoop启动过程是怎样的”的有关知识,在实际案例的操作过程中,不少人都会遇到这样的困境,接下来就让小编带领大家学习一下如何处理这些情况吧!希望大家仔细阅读,能够学有所成!
分析NioEventLoop的execute()接口,主要逻辑如下:
添加任务队列
绑定当前线程到EventLoop上
调用EventLoop的run()方法
private static void doBind0( final ChannelFuture regFuture, final Channel channel, final SocketAddress localAddress, final ChannelPromise promise) { // 通过eventLoop来执行channel绑定的Task channel.eventLoop().execute(new Runnable() { @Override public void run() { if (regFuture.isSuccess()) { // channel绑定 channel.bind(localAddress, promise).addListener(ChannelFutureListener.CLOSE_ON_FAILURE); } else { promise.setFailure(regFuture.cause()); } } }); }
往下追踪到 SingleThreadEventExecutor 中 execute 接口,如下:
@Overridepublic void execute(Runnable task) { if (task == null) { throw new NullPointerException("task"); } // 判断当前运行时线程是否与EventLoop中绑定的线程一致 // 这里还未绑定Thread,所以先返回false boolean inEventLoop = inEventLoop(); // 将任务添加任务队列,也就是我们前面讲EventLoop创建时候提到的 MpscQueue. addTask(task); if (!inEventLoop) { // 启动线程 startThread(); if (isShutdown() && removeTask(task)) { reject(); } } if (!addTaskWakesUp && wakesUpForTask(task)) { wakeup(inEventLoop); } }
启动线程接口:
private void startThread() { // 状态比较,最开始时state = 1 ,为true if (state == ST_NOT_STARTED) { // cs操作后,state状态设置为 2 if (STATE_UPDATER.compareAndSet(this, ST_NOT_STARTED, ST_STARTED)) { try { // 启动接口 doStartThread(); } catch (Throwable cause) { STATE_UPDATER.set(this, ST_NOT_STARTED); PlatformDependent.throwException(cause); } } } }// 执行线程启动方法private void doStartThread() { // 断言判断 SingleThreadEventExecutor 还未绑定 Thread assert thread == null; // executor 执行任务 executor.execute(new Runnable() { @Override public void run() { // 将 SingleThreadEventExecutor(在我们的案例中就是NioEventLoop) 与 当前线程进行绑定 thread = Thread.currentThread(); if (interrupted) { thread.interrupt(); } // 设置状态为 false boolean success = false; // 更新最近一次任务的执行时间 updateLastExecutionTime(); try { // 往下调用 NioEventLoop 的 run 方法,执行 SingleThreadEventExecutor.this.run(); success = true; } catch (Throwable t) { logger.warn("Unexpected exception from an event executor: ", t); } finally { ... } } }); }
往下调用到 NioEventLoop 中的 run 方法,通过无限for循环,主要做以下三件事情:
轮循I/O事件:select(wakenUp.getAndSet(false))
处理I/O事件:processSelectedKeys
运行Task任务:runAllTasks
@Overrideprotected void run() { for (;;) { try { switch (selectStrategy.calculateStrategy(selectNowSupplier, hasTasks())) { case SelectStrategy.CONTINUE: continue; case SelectStrategy.SELECT: // 轮训检测I/O事件 // wakenUp为了标记selector是否是唤醒状态,每次select操作,都设置为false,也就是未唤醒状态。 select(wakenUp.getAndSet(false)); // 'wakenUp.compareAndSet(false, true)' 总是在调用 'selector.wakeup()' 之前进行评估,以减少唤醒的开销 // (Selector.wakeup() 是非常耗性能的操作.) // 但是,这种方法存在竞争条件。当「wakeup」太早设置为true时触发竞争条件 // 在下面两种情况下,「wakenUp」会过早设置为true: // 1)Selector 在 'wakenUp.set(false)' 与 'selector.select(...)' 之间被唤醒。(BAD) // 2)Selector 在 'selector.select(...)' 与 'if (wakenUp.get()) { ... }' 之间被唤醒。(OK) // 在第一种情况下,'wakenUp'设置为true,后面的'selector.select(...)'将立即唤醒。 直到'wakenUp'在下一轮中再次设置为false,'wakenUp.compareAndSet(false,true)'将失败,因此任何唤醒选择器的尝试也将失败,从而导致以下'selector.select(。 ..)'呼吁阻止不必要的。 // 要解决这个问题,如果在selector.select(...)操作之后wakenUp立即为true,我们会再次唤醒selector。 它是低效率的,因为它唤醒了第一种情况(BAD - 需要唤醒)和第二种情况(OK - 不需要唤醒)的选择器。 if (wakenUp.get()) { selector.wakeup(); } // fall through default: } cancelledKeys = 0; needsToSelectAgain = false; // ioRatio 表示处理I/O事件与执行具体任务事件之间所耗时间的比值。 // ioRatio 默认为50 final int ioRatio = this.ioRatio; if (ioRatio == 100) { try { // 处理I/O事件 processSelectedKeys(); } finally { // 处理任务队列 runAllTasks(); } } else { // 处理IO事件的开始时间 final long ioStartTime = System.nanoTime(); try { // 处理I/O事件 processSelectedKeys(); } finally { // 记录io所耗时间 final long ioTime = System.nanoTime() - ioStartTime; // 处理任务队列,设置最大的超时时间 runAllTasks(ioTime * (100 - ioRatio) / ioRatio); } } } catch (Throwable t) { handleLoopException(t); } // Always handle shutdown even if the loop processing threw an exception. try { if (isShuttingDown()) { closeAll(); if (confirmShutdown()) { return; } } } catch (Throwable t) { handleLoopException(t); } } }
private void select(boolean oldWakenUp) throws IOException { Selector selector = this.selector; try { // select操作计数 int selectCnt = 0; // 记录当前系统时间 long currentTimeNanos = System.nanoTime(); // delayNanos方法用于计算定时任务队列,最近一个任务的截止时间 // selectDeadLineNanos 表示当前select操作所不能超过的最大截止时间 long selectDeadLineNanos = currentTimeNanos + delayNanos(currentTimeNanos); for (;;) { // 计算超时时间,判断是否超时 long timeoutMillis = (selectDeadLineNanos - currentTimeNanos + 500000L) / 1000000L; // 如果 timeoutMillis <= 0, 表示超时,进行一个非阻塞的 select 操作。设置 selectCnt 为 1. 并终止本次循环。 if (timeoutMillis <= 0) { if (selectCnt == 0) { selector.selectNow(); selectCnt = 1; } break; } // 当wakenUp为ture时,恰好有task被提交,这个task将无法获得调用的机会 // Selector#wakeup. 因此,在执行select操作之前,需要再次检查任务队列 // 如果不这么做,这个Task将一直挂起,直到select操作超时 // 如果 pipeline 中存在 IdleStateHandler ,那么Task将一直挂起直到 空闲超时。 if (hasTasks() && wakenUp.compareAndSet(false, true)) { // 调用非阻塞方法 selector.selectNow(); selectCnt = 1; break; } // 如果当前任务队列为空,并且超时时间未到,则进行一个阻塞式的selector操作。timeoutMillis 为最大的select时间 int selectedKeys = selector.select(timeoutMillis); // 操作计数 +1 selectCnt ++; // 存在以下情况,本次selector则终止 if (selectedKeys != 0 || oldWakenUp || wakenUp.get() || hasTasks() || hasScheduledTasks()) { // - 轮训到了事件(Selected something,) // - 被用户唤醒(waken up by user,) // - 已有任务队列(the task queue has a pending task.) // - 已有定时任务(a scheduled task is ready for processing) break; } if (Thread.interrupted()) { // Thread was interrupted so reset selected keys and break so we not run into a busy loop. // As this is most likely a bug in the handler of the user or it's client library we will // also log it. // // See https://github.com/netty/netty/issues/2426 if (logger.isDebugEnabled()) { logger.debug("Selector.select() returned prematurely because " + "Thread.currentThread().interrupt() was called. Use " + "NioEventLoop.shutdownGracefully() to shutdown the NioEventLoop."); } selectCnt = 1; break; } // 记录当前时间 long time = System.nanoTime(); // 如果time > currentTimeNanos + timeoutMillis(超时时间),则表明已经执行过一次select操作 if (time - TimeUnit.MILLISECONDS.toNanos(timeoutMillis) >= currentTimeNanos) { // timeoutMillis elapsed without anything selected. selectCnt = 1; } // 如果 time <= currentTimeNanos + timeoutMillis,表示触发了空轮训 // 如果空轮训的次数超过 SELECTOR_AUTO_REBUILD_THRESHOLD (512),则重建一个新的selctor,避免空轮训 else if (SELECTOR_AUTO_REBUILD_THRESHOLD > 0 && selectCnt >= SELECTOR_AUTO_REBUILD_THRESHOLD) { // The selector returned prematurely many times in a row. // Rebuild the selector to work around the problem. logger.warn( "Selector.select() returned prematurely {} times in a row; rebuilding Selector {}.", selectCnt, selector); // 重建创建一个新的selector rebuildSelector(); selector = this.selector; // Select again to populate selectedKeys. // 对重建后的selector进行一次非阻塞调用,用于获取最新的selectedKeys selector.selectNow(); // 设置select计数 selectCnt = 1; break; } currentTimeNanos = time; } if (selectCnt > MIN_PREMATURE_SELECTOR_RETURNS) { if (logger.isDebugEnabled()) { logger.debug("Selector.select() returned prematurely {} times in a row for Selector {}.", selectCnt - 1, selector); } } } catch (CancelledKeyException e) { if (logger.isDebugEnabled()) { logger.debug(CancelledKeyException.class.getSimpleName() + " raised by a Selector {} - JDK bug?", selector, e); } // Harmless exception - log anyway } }
该方法的主要逻辑就是:
创建一个新的selector
将老的selector上的 selectKey注册到新的 selector 上
public void rebuildSelector() { if (!inEventLoop()) { execute(new Runnable() { @Override public void run() { rebuildSelector0(); } }); return; } rebuildSelector0(); }// 重新创建selectorprivate void rebuildSelector0() { // 暂存老的selector final Selector oldSelector = selector; final SelectorTuple newSelectorTuple; if (oldSelector == null) { return; } try { // 创建一个新的 SelectorTuple // openSelector()在之前分析过了 newSelectorTuple = openSelector(); } catch (Exception e) { logger.warn("Failed to create a new Selector.", e); return; } // Register all channels to the new Selector. // 记录select上注册的channel数量 int nChannels = 0; // 遍历老的 selector 上的 SelectionKey for (SelectionKey key: oldSelector.keys()) { // 获取 attachment,这里的attachment就是我们前面在讲 Netty Channel注册时,select会将channel赋值到 attachment 变量上。 // 获取老的selector上注册的channel Object a = key.attachment(); try { if (!key.isValid() || key.channel().keyFor(newSelectorTuple.unwrappedSelector) != null) { continue; } // 获取兴趣集 int interestOps = key.interestOps(); // 取消 SelectionKey key.cancel(); // 将老的兴趣集重新注册到前面新创建的selector上 SelectionKey newKey = key.channel().register(newSelectorTuple.unwrappedSelector, interestOps, a); if (a instanceof AbstractNioChannel) { // Update SelectionKey ((AbstractNioChannel) a).selectionKey = newKey; } // nChannels计数 + 1 nChannels ++; } catch (Exception e) { logger.warn("Failed to re-register a Channel to the new Selector.", e); if (a instanceof AbstractNioChannel) { AbstractNioChannel ch = (AbstractNioChannel) a; ch.unsafe().close(ch.unsafe().voidPromise()); } else { @SuppressWarnings("unchecked") NioTask<SelectableChannel> task = (NioTask<SelectableChannel>) a; invokeChannelUnregistered(task, key, e); } } } // 设置新的 selector selector = newSelectorTuple.selector; // 设置新的 unwrappedSelector unwrappedSelector = newSelectorTuple.unwrappedSelector; try { // time to close the old selector as everything else is registered to the new one // 关闭老的seleclor oldSelector.close(); } catch (Throwable t) { if (logger.isWarnEnabled()) { logger.warn("Failed to close the old Selector.", t); } } if (logger.isInfoEnabled()) { logger.info("Migrated " + nChannels + " channel(s) to the new Selector."); } }
private void processSelectedKeysOptimized() { for (int i = 0; i < selectedKeys.size; ++i) { final SelectionKey k = selectedKeys.keys[i]; // null out entry in the array to allow to have it GC'ed once the Channel close // See https://github.com/netty/netty/issues/2363 // 设置为null,有利于GC回收 selectedKeys.keys[i] = null; // 获取 SelectionKey 中的 attachment, 我们这里就是 NioChannel final Object a = k.attachment(); if (a instanceof AbstractNioChannel) { // 处理 SelectedKey processSelectedKey(k, (AbstractNioChannel) a); } else { @SuppressWarnings("unchecked") NioTask<SelectableChannel> task = (NioTask<SelectableChannel>) a; processSelectedKey(k, task); } if (needsToSelectAgain) { // null out entries in the array to allow to have it GC'ed once the Channel close // See https://github.com/netty/netty/issues/2363 selectedKeys.reset(i + 1); selectAgain(); i = -1; } } }// 处理 SelectedKeyprivate void processSelectedKey(SelectionKey k, AbstractNioChannel ch) { // 获取Netty Channel中的 NioUnsafe 对象,用于后面的IO操作 final AbstractNioChannel.NioUnsafe unsafe = ch.unsafe(); // 判断 SelectedKey 的有效性,如果无效,则直接返回并关闭channel if (!k.isValid()) { final EventLoop eventLoop; try { eventLoop = ch.eventLoop(); } catch (Throwable ignored) { // If the channel implementation throws an exception because there is no event loop, we ignore this // because we are only trying to determine if ch is registered to this event loop and thus has authority // to close ch. return; } // Only close ch if ch is still registered to this EventLoop. ch could have deregistered from the event loop // and thus the SelectionKey could be cancelled as part of the deregistration process, but the channel is // still healthy and should not be closed. // See https://github.com/netty/netty/issues/5125 if (eventLoop != this || eventLoop == null) { return; } // close the channel if the key is not valid anymore // 关闭channel unsafe.close(unsafe.voidPromise()); return; } try { // 获取 SelectionKey 中所有准备就绪的操作集 int readyOps = k.readyOps(); // We first need to call finishConnect() before try to trigger a read(...) or write(...) as otherwise // the NIO JDK channel implementation may throw a NotYetConnectedException. // 在调用处理READ与WRITE事件之间,先调用finishConnect()接口,避免异常 NotYetConnectedException 发生。 if ((readyOps & SelectionKey.OP_CONNECT) != 0) { // remove OP_CONNECT as otherwise Selector.select(..) will always return without blocking // See https://github.com/netty/netty/issues/924 int ops = k.interestOps(); ops &= ~SelectionKey.OP_CONNECT; k.interestOps(ops); unsafe.finishConnect(); } // Process OP_WRITE first as we may be able to write some queued buffers and so free memory. // 处理 WRITE 事件 if ((readyOps & SelectionKey.OP_WRITE) != 0) { // Call forceFlush which will also take care of clear the OP_WRITE once there is nothing left to write ch.unsafe().forceFlush(); } // Also check for readOps of 0 to workaround possible JDK bug which may otherwise lead // to a spin loop // 处理 ACCEPT 与 READ 事件 // 如果当前的EventLoop是WorkGroup,则表示有 READ 事件 // 如果当前的EventLoop是BossGroup,则表示有 ACCEPT 事件,有新连接进来了 if ((readyOps & (SelectionKey.OP_READ | SelectionKey.OP_ACCEPT)) != 0 || readyOps == 0) { // 读取数据 unsafe.read(); } } catch (CancelledKeyException ignored) { unsafe.close(unsafe.voidPromise()); } }
关于
unsafe.read()
的分析,请看 后文
接下来,我们了解一下执行具体Task任务的接口:runAllTasks。在EventLoop中,待执行的任务队列分为两种:一种是普通任务队列,一种是定时任务队列。
前面 我们讲 EventLoop 创建时提到过NioEventLoop中 taskQueue 的创建,是一个MpscQueue,关于高效率的MpscQueue 后面单独写文章进行介绍:
public abstract class SingleThreadEventExecutor extends AbstractScheduledEventExecutor implements OrderedEventExecutor { ... // 存放普通任务的队列 private final Queue<Runnable> taskQueue; ... protected SingleThreadEventExecutor(EventExecutorGroup parent, Executor executor, boolean addTaskWakesUp, int maxPendingTasks, RejectedExecutionHandler rejectedHandler) { super(parent); this.addTaskWakesUp = addTaskWakesUp; this.maxPendingTasks = Math.max(16, maxPendingTasks); this.executor = ObjectUtil.checkNotNull(executor, "executor"); // 创建TaskQueue taskQueue = newTaskQueue(this.maxPendingTasks); rejectedExecutionHandler = ObjectUtil.checkNotNull(rejectedHandler, "rejectedHandler"); } ... }public final class NioEventLoop extends SingleThreadEventLoop { ... // NioEventLoop 创建TaskQueue队列 @Override protected Queue<Runnable> newTaskQueue(int maxPendingTasks) { // This event loop never calls takeTask() return maxPendingTasks == Integer.MAX_VALUE ? PlatformDependent.<Runnable>newMpscQueue() : PlatformDependent.<Runnable>newMpscQueue(maxPendingTasks); } ... }
存放定时任务的队列在 AbstractScheduledEventExecutor 中,成员变量为 scheduledTaskQueue,代码如下:
public abstract class AbstractScheduledEventExecutor extends AbstractEventExecutor { // 优先级队列的比较器 private static final Comparator<ScheduledFutureTask<?>> SCHEDULED_FUTURE_TASK_COMPARATOR = new Comparator<ScheduledFutureTask<?>>() { @Override public int compare(ScheduledFutureTask<?> o1, ScheduledFutureTask<?> o2) { return o1.compareTo(o2); } }; // 存放定时任务的优先级队列 PriorityQueue<ScheduledFutureTask<?>> scheduledTaskQueue; // 创建定时任务队列 PriorityQueue<ScheduledFutureTask<?>> scheduledTaskQueue() { if (scheduledTaskQueue == null) { scheduledTaskQueue = new DefaultPriorityQueue<ScheduledFutureTask<?>>( SCHEDULED_FUTURE_TASK_COMPARATOR, // Use same initial capacity as java.util.PriorityQueue 11); } return scheduledTaskQueue; } // 保存定时任务 @Override public ScheduledFuture<?> schedule(Runnable command, long delay, TimeUnit unit) { ObjectUtil.checkNotNull(command, "command"); ObjectUtil.checkNotNull(unit, "unit"); if (delay < 0) { delay = 0; } validateScheduled0(delay, unit); return schedule(new ScheduledFutureTask<Void>( this, command, null, ScheduledFutureTask.deadlineNanos(unit.toNanos(delay)))); } // 保存定时任务 @Override public <V> ScheduledFuture<V> schedule(Callable<V> callable, long delay, TimeUnit unit) { ObjectUtil.checkNotNull(callable, "callable"); ObjectUtil.checkNotNull(unit, "unit"); if (delay < 0) { delay = 0; } validateScheduled0(delay, unit); return schedule(new ScheduledFutureTask<V>( this, callable, ScheduledFutureTask.deadlineNanos(unit.toNanos(delay)))); } // 保存定时任务 <V> ScheduledFuture<V> schedule(final ScheduledFutureTask<V> task) { // 判断是否为当前线程 if (inEventLoop()) { // 添加定时任务队列 scheduledTaskQueue().add(task); } else { execute(new Runnable() { @Override public void run() { // 添加定时任务队列 scheduledTaskQueue().add(task); } }); } return task; } }
Netty存放定时任务队列为 DefaultPriorityQueue ,定时任务的封装对象为 ScheduledFutureTask ,在队列中的优先按照它们的截止时间进行排序,其次在按照id进行排序。
final class ScheduledFutureTask<V> extends PromiseTask<V> implements ScheduledFuture<V>, PriorityQueueNode { ... // 比较 ScheduledFutureTask 之间的排序 @Override public int compareTo(Delayed o) { if (this == o) { return 0; } ScheduledFutureTask<?> that = (ScheduledFutureTask<?>) o; long d = deadlineNanos() - that.deadlineNanos(); if (d < 0) { return -1; } else if (d > 0) { return 1; } else if (id < that.id) { return -1; } else if (id == that.id) { throw new Error(); } else { return 1; } } ... }
再来看看任务的执行逻辑,首先将定时任务取出,聚合到普通任务队列中,再去for循环运行每个Task。
protected boolean runAllTasks(long timeoutNanos) { // 将定时任务从定时队列中取出,放入普通队列中 fetchFromScheduledTaskQueue(); // 从队列中取出任务 Runnable task = pollTask(); if (task == null) { afterRunningAllTasks(); return false; } // 计算任务执行的最大超时时间 final long deadline = ScheduledFutureTask.nanoTime() + timeoutNanos; // 任务计数 long runTasks = 0; // 最近一次任务执行的时间 long lastExecutionTime; for (;;) { // 执行任务 safeExecute(task); // 任务计数 +1 runTasks ++; // Check timeout every 64 tasks because nanoTime() is relatively expensive. // XXX: Hard-coded value - will make it configurable if it is really a problem. // 由于nanoTime() 是非常好性能的操作,因此每64次就对比一下 定时任务的执行时间与 deadline, // 如果 lastExecutionTime >= deadline,则表示任务超时了,需要中断退出 if ((runTasks & 0x3F) == 0) { lastExecutionTime = ScheduledFutureTask.nanoTime(); if (lastExecutionTime >= deadline) { break; } } // 获取任务 task = pollTask(); if (task == null) { lastExecutionTime = ScheduledFutureTask.nanoTime(); break; } } afterRunningAllTasks(); // 记录最后一次的执行时间 this.lastExecutionTime = lastExecutionTime; return true; }// 取出任务protected Runnable pollTask() { assert inEventLoop(); return pollTaskFrom(taskQueue); }// 从队列中取出任务protected static Runnable pollTaskFrom(Queue<Runnable> taskQueue) { for (;;) { Runnable task = taskQueue.poll(); if (task == WAKEUP_TASK) { continue; } return task; } }// 将定时任务从定时队列中取出,聚合到普通队列中:private boolean fetchFromScheduledTaskQueue() { // 得到nanoTime = 当前时间 - ScheduledFutureTask的START_TIME(开始时间) long nanoTime = AbstractScheduledEventExecutor.nanoTime(); // 获得截止时间小于nanoTime的定时任务 Runnable scheduledTask = pollScheduledTask(nanoTime); while (scheduledTask != null) { // 将定时任务放入普通队列中,以备运行 if (!taskQueue.offer(scheduledTask)) { // No space left in the task queue add it back to the scheduledTaskQueue so we pick it up again. // 如果 taskQueue 没有足够的空间,导致添加失败,则将其返回定时任务队列中 scheduledTaskQueue().add((ScheduledFutureTask<?>) scheduledTask); return false; } scheduledTask = pollScheduledTask(nanoTime); } return true; }// 获得截止时间小于nanoTime的定时任务protected final Runnable pollScheduledTask(long nanoTime) { assert inEventLoop(); // 获取定时任务队列 Queue<ScheduledFutureTask<?>> scheduledTaskQueue = this.scheduledTaskQueue; // 获取第一个定时任务 ScheduledFutureTask<?> scheduledTask = scheduledTaskQueue == null ? null : scheduledTaskQueue.peek(); if (scheduledTask == null) { return null; } // 如果该定时任务的截止时间 <= nanoTime ,则返回 if (scheduledTask.deadlineNanos() <= nanoTime) { scheduledTaskQueue.remove(); return scheduledTask; } return null; }
“Netty NioEventLoop启动过程是怎样的”的内容就介绍到这里了,感谢大家的阅读。如果想了解更多行业相关的知识可以关注亿速云网站,小编将为大家输出更多高质量的实用文章!
免责声明:本站发布的内容(图片、视频和文字)以原创、转载和分享为主,文章观点不代表本网站立场,如果涉及侵权请联系站长邮箱:is@yisu.com进行举报,并提供相关证据,一经查实,将立刻删除涉嫌侵权内容。