Elasticsearch 6.1.0线程池介绍

  1. 云栖社区>
  2. 博客列表>
  3. 正文

Elasticsearch 6.1.0线程池介绍

晴天哥 2018-12-07 17:47:15 浏览166 评论0

摘要: 开篇  这篇文章主要是用来讲解ES线程池(EsExecutors)的实现,然后象征性的和JDK的Executors实现进行了简单的对比,看了这篇文章以后要对Executors和ThreadPoolExecutor的使用更有信心才对。

开篇

 这篇文章主要是用来讲解ES线程池(EsExecutors)的实现,然后象征性的和JDK的Executors实现进行了简单的对比,看了这篇文章以后要对Executors和ThreadPoolExecutor的使用更有信心才对。


elasticsearch线程池配置

public class ThreadPool extends AbstractComponent implements Scheduler, Closeable {

  final int availableProcessors = EsExecutors.numberOfProcessors(settings);
  final int halfProcMaxAt5 = halfNumberOfProcessorsMaxFive(availableProcessors);
  final int halfProcMaxAt10 = halfNumberOfProcessorsMaxTen(availableProcessors);
  final int genericThreadPoolMax = boundedBy(4 * availableProcessors, 128, 512);

  builders.put(Names.GENERIC, 
  new ScalingExecutorBuilder(Names.GENERIC, 4, genericThreadPoolMax, TimeValue.timeValueSeconds(30)));

  builders.put(Names.INDEX, 
  new FixedExecutorBuilder(settings, Names.INDEX, availableProcessors, 200));

  builders.put(Names.BULK, 
  new FixedExecutorBuilder(settings, Names.BULK, availableProcessors, 200)); 

  builders.put(Names.GET, 
  new FixedExecutorBuilder(settings, Names.GET, availableProcessors, 1000));

  builders.put(Names.SEARCH, 
  new AutoQueueAdjustingExecutorBuilder(settings,
                        Names.SEARCH, searchThreadPoolSize(availableProcessors), 1000, 1000, 1000, 2000));

  builders.put(Names.MANAGEMENT, 
  new ScalingExecutorBuilder(Names.MANAGEMENT, 1, 5, TimeValue.timeValueMinutes(5)));

  builders.put(Names.LISTENER, 
  new FixedExecutorBuilder(settings, Names.LISTENER, halfProcMaxAt10, -1));

  builders.put(Names.FLUSH, 
  new ScalingExecutorBuilder(Names.FLUSH, 1, halfProcMaxAt5, TimeValue.timeValueMinutes(5)));

  builders.put(Names.REFRESH, 
  new ScalingExecutorBuilder(Names.REFRESH, 1, halfProcMaxAt10, TimeValue.timeValueMinutes(5)));

  builders.put(Names.WARMER, 
  new ScalingExecutorBuilder(Names.WARMER, 1, halfProcMaxAt5, TimeValue.timeValueMinutes(5)));

  builders.put(Names.SNAPSHOT, 
  new ScalingExecutorBuilder(Names.SNAPSHOT, 1, halfProcMaxAt5, TimeValue.timeValueMinutes(5)));

  builders.put(Names.FETCH_SHARD_STARTED, 
  new ScalingExecutorBuilder(Names.FETCH_SHARD_STARTED, 1, 2 * availableProcessors, TimeValue.timeValueMinutes(5)));

  builders.put(Names.FORCE_MERGE, 
  new FixedExecutorBuilder(settings, Names.FORCE_MERGE, 1, -1));

  builders.put(Names.FETCH_SHARD_STORE, 
  new ScalingExecutorBuilder(Names.FETCH_SHARD_STORE, 1, 2 * availableProcessors, TimeValue.timeValueMinutes(5)));
}

说明:

  • elasticsearch线程池根据作用的不同主要分为两大类 ScalingExecutor和FixedExecutor。
  • ScalingExecutor表示线程池中的线程数是动态可变的。
  • FixedExecutor表示线程池中的线程池是不可变的。


elasticsearch线程池分类

elasticsearch线程池的线程按照源码的实现来看分为FIXED和SCALING两大类FIXED的意思是固定线程的数量(core thread个数 = max thread个数),SCALING的意思是动态调整线程数量(core thread个数 != max thread个数)。

FIXED

说明:大小固定设置的threadpool,它有一个queue来存放pending的请求,其中pool的大小默认是core*5,queue_size默认是-1(即是无限制)。

  • LISTENER:用作client的操作,默认大小halfProcMaxAt10,queue_size=-1无限制;
  • GET:用作get操作,默认大小availableProcessors,queue_size为1000;
  • INDEX:用作index或delete操作,默认大小availableProcessors,queue_size为200;
  • BULK:用作bulk操作,默认大小为availableProcessors,queue_size为200;
  • SEARCH:用作count或是search操作,默认大小((availableProcessors * 3) / 2) + 1;queue_size为1000;
  • SUGGEST:用作suggest操作,默认大小availableProcessors,queue_size为1000;
  • PERCOLATE:用作percolate,默认大小为availableProcessors,queue_size为1000;
  • FORCE_MERGE:用作force_merge操作(2.1之前叫做optimize),默认大小为1;


SCALING

说明:拥有可变大小的pool,其值可在1和设置值之间。

  • GENERIC:通用的操作,比如node的discovery,默认大小genericThreadPoolMax,默认keep alive时间是30sec;
  • MANAGEMENT:用作ES的管理,比如集群的管理;默认大小5,keep alive时间为5min;
  • FLUSH:用作flush操作,默认大小为halfProcMaxAt5,keep alive时间为5min;
  • REFRESH:用作refresh操作,默认大小为halfProcMaxAt10,keep alive时间为5min;
  • WARMER:用作index warm-up操作,默认大小为halfProcMaxAt5,keep alive时间为5min;
  • SNAPSHOT:用作snapshot操作,默认大小为halfProcMaxAt5,keep alive时间为5min;
  • FETCH_SHARD_STARTED:用作fetch shard开始操作,默认大小availableProcessors * 2,keep alive时间为5min;
  • FETCH_SHARD_STORE:用作fetch shard存储操作,默认大小availableProcessors * 2,keep alive时间为5min;


JDK的Executors

public class Executors {

    public static ExecutorService newFixedThreadPool(int nThreads) {
        return new ThreadPoolExecutor(nThreads, nThreads,
                                      0L, TimeUnit.MILLISECONDS,
                                      new LinkedBlockingQueue<Runnable>());
    }

    public static ExecutorService newFixedThreadPool(int nThreads, ThreadFactory threadFactory) {
        return new ThreadPoolExecutor(nThreads, nThreads,
                                      0L, TimeUnit.MILLISECONDS,
                                      new LinkedBlockingQueue<Runnable>(),
                                      threadFactory);
    }

    public static ExecutorService newCachedThreadPool() {
        return new ThreadPoolExecutor(0, Integer.MAX_VALUE,
                                      60L, TimeUnit.SECONDS,
                                      new SynchronousQueue<Runnable>());
    }

    public static ExecutorService newCachedThreadPool(ThreadFactory threadFactory) {
        return new ThreadPoolExecutor(0, Integer.MAX_VALUE,
                                      60L, TimeUnit.SECONDS,
                                      new SynchronousQueue<Runnable>(),
                                      threadFactory);
    }
}

说明:

  • Executors的newFixedThreadPool创建固定线程数量的线程池。
  • Executors的newCachedThreadPool创建可动态调整线程数量的线程池。
  • Executors创建的是ThreadPoolExecutor对象。


Elasticsearch的EsExecutors

public class EsExecutors {

    public static final Setting<Integer> PROCESSORS_SETTING =
        Setting.intSetting("processors", Runtime.getRuntime().availableProcessors(), 1, Property.NodeScope);

    public static int numberOfProcessors(final Settings settings) {
        return PROCESSORS_SETTING.get(settings);
    }

    public static EsThreadPoolExecutor newScaling(String name, int min, int max, long keepAliveTime, 
                          TimeUnit unit, ThreadFactory threadFactory, ThreadContext contextHolder) {
        ExecutorScalingQueue<Runnable> queue = new ExecutorScalingQueue<>();
        EsThreadPoolExecutor executor = new EsThreadPoolExecutor(name, min, max, keepAliveTime, 
                            unit, queue, threadFactory, new ForceQueuePolicy(), contextHolder);
        queue.executor = executor;
        return executor;
    }

    public static EsThreadPoolExecutor newFixed(String name, int size, int queueCapacity, 
                             ThreadFactory threadFactory, ThreadContext contextHolder) {
        BlockingQueue<Runnable> queue;
        if (queueCapacity < 0) {
            queue = ConcurrentCollections.newBlockingQueue();
        } else {
            queue = new SizeBlockingQueue<>(ConcurrentCollections.<Runnable>newBlockingQueue(), queueCapacity);
        }
        return new EsThreadPoolExecutor(name, size, size, 0, TimeUnit.MILLISECONDS, queue, 
                                        threadFactory, new EsAbortPolicy(), contextHolder);
    }

    public static EsThreadPoolExecutor newAutoQueueFixed(String name, int size, int initialQueueCapacity, int minQueueSize,
                                                         int maxQueueSize, int frameSize, TimeValue targetedResponseTime,
                                                         ThreadFactory threadFactory, ThreadContext contextHolder) {
        if (initialQueueCapacity <= 0) {
            throw new IllegalArgumentException("initial queue capacity for [" + name + "] executor must be positive, got: " +
                            initialQueueCapacity);
        }
        ResizableBlockingQueue<Runnable> queue =
                new ResizableBlockingQueue<>(ConcurrentCollections.<Runnable>newBlockingQueue(), initialQueueCapacity);
        return new QueueResizingEsThreadPoolExecutor(name, size, size, 0, TimeUnit.MILLISECONDS,
                queue, minQueueSize, maxQueueSize, TimedRunnable::new, frameSize, targetedResponseTime, threadFactory,
                new EsAbortPolicy(), contextHolder);
    }
}

说明:

  • newScaling()方法创建可扩展线程数量的线程池,淘汰策略使用ForceQueuePolicy。
  • newFixed()方法创建创建固定线程数量的线程池,淘汰策略使用EsAbortPolicy。
  • newAutoQueueFixed()方法创建固定线程数量但是Queue队列数量可以动态调整的线程池,淘汰策略使用EsAbortPolicy。。
  • EsExecutors内部创建的是EsThreadPoolExecutor对象。
  • EsExecutors的实现借鉴了JDK的Executors接口,给我们提供了自定Executors的思路。


EsExecutors的EsThreadPoolExecutor

public class EsThreadPoolExecutor extends ThreadPoolExecutor {
    EsThreadPoolExecutor(String name, int corePoolSize, int maximumPoolSize, long keepAliveTime, TimeUnit unit,
            BlockingQueue<Runnable> workQueue, ThreadFactory threadFactory, ThreadContext contextHolder) {
        this(name, corePoolSize, maximumPoolSize, keepAliveTime, unit, 
             workQueue, threadFactory, new EsAbortPolicy(), contextHolder);
    }

    EsThreadPoolExecutor(String name, int corePoolSize, int maximumPoolSize, long keepAliveTime, TimeUnit unit,
            BlockingQueue<Runnable> workQueue, ThreadFactory threadFactory, XRejectedExecutionHandler handler,
            ThreadContext contextHolder) {
        super(corePoolSize, maximumPoolSize, keepAliveTime, unit, workQueue, threadFactory, handler);
        this.name = name;
        this.contextHolder = contextHolder;
    }

    @Override
    public void execute(final Runnable command) {
        doExecute(wrapRunnable(command));
    }

    protected void doExecute(final Runnable command) {
        try {
            super.execute(command);
        } catch (EsRejectedExecutionException ex) {
            if (command instanceof AbstractRunnable) {
                try {
                    ((AbstractRunnable) command).onRejection(ex);
                } finally {
                    ((AbstractRunnable) command).onAfter();

                }
            } else {
                throw ex;
            }
        }
    }

    @Override
    protected void afterExecute(Runnable r, Throwable t) {
        super.afterExecute(r, t);
        assert assertDefaultContext(r);
    }
}

说明:

  • EsThreadPoolExecutor继承自ThreadPoolExecutor对象,构造函数内部初始化ThreadPoolExecutor对象。
  • EsThreadPoolExecutor的核心的execute方法内部也是调用了ThreadPoolExecutor的execute方法。
  • EsThreadPoolExecutor重新了execute和afterExecute方法。
  • EsThreadPoolExecutor给我们提供了重写ThreadPoolExecutor的思路,值得学习。


EsExecutors的ThreadFactory

public class EsExecutors {

    public static String threadName(Settings settings, String ... names) {
        String namePrefix =
                Arrays
                        .stream(names)
                        .filter(name -> name != null)
                        .collect(Collectors.joining(".", "[", "]"));
        return threadName(settings, namePrefix);
    }

    public static ThreadFactory daemonThreadFactory(String namePrefix) {
        return new EsThreadFactory(namePrefix);
    }

    static class EsThreadFactory implements ThreadFactory {

        final ThreadGroup group;
        final AtomicInteger threadNumber = new AtomicInteger(1);
        final String namePrefix;

        EsThreadFactory(String namePrefix) {
            this.namePrefix = namePrefix;
            SecurityManager s = System.getSecurityManager();
            group = (s != null) ? s.getThreadGroup() :
                    Thread.currentThread().getThreadGroup();
        }

        @Override
        public Thread newThread(Runnable r) {
            Thread t = new Thread(group, r,
                    namePrefix + "[T#" + threadNumber.getAndIncrement() + "]",
                    0);
            t.setDaemon(true);
            return t;
        }
    }

    private EsExecutors() {
    }
}

说明:

  • 1.EsExecutors给我们提供一种创建线程工厂的标准方法,实现ThreadFactory接口重新newThread()方法。
  • 2.通过AtomicInteger threadNumber = new AtomicInteger(1)变量生成线程自增的线程id。
  • 3.线程池的线程都有具体意义的线程名非常重要有利于排查问题,非常推荐使用。


EsExecutors的AbortPolicy

public interface XRejectedExecutionHandler extends RejectedExecutionHandler {
    long rejected();
}

public class EsAbortPolicy implements XRejectedExecutionHandler {
    private final CounterMetric rejected = new CounterMetric();

    @Override
    public void rejectedExecution(Runnable r, ThreadPoolExecutor executor) {
        if (r instanceof AbstractRunnable) {
            if (((AbstractRunnable) r).isForceExecution()) {
                BlockingQueue<Runnable> queue = executor.getQueue();
                if (!(queue instanceof SizeBlockingQueue)) {
                    throw new IllegalStateException("forced execution, but expected a size queue");
                }
                try {
                    ((SizeBlockingQueue) queue).forcePut(r);
                } catch (InterruptedException e) {
                    Thread.currentThread().interrupt();
                    throw new IllegalStateException("forced execution, but got interrupted", e);
                }
                return;
            }
        }
        rejected.inc();
        throw new EsRejectedExecutionException("rejected execution of " + r + " on " + executor, executor.isShutdown());
    }

    public long rejected() {
        return rejected.count();
    }
}


static class ForceQueuePolicy implements XRejectedExecutionHandler {
        @Override
        public void rejectedExecution(Runnable r, ThreadPoolExecutor executor) {
            try {
                executor.getQueue().put(r);
            } catch (InterruptedException e) {
                throw new EsRejectedExecutionException(e);
            }
        }

        @Override
        public long rejected() {
            return 0;
        }
    }

说明:

  • EsAbortPolicy的过期策略提供了我们自定实现过期策略的案例。


参考文章

Elasticsearch源码3(线程池)

【云栖快讯】阿里云栖开发者沙龙(Java技术专场)火热来袭!快来报名参与吧!  详情请点击

网友评论