美文网首页
重写并优化 ThreadPoolExecutor 线程池

重写并优化 ThreadPoolExecutor 线程池

作者: 备货仓66 | 来源:发表于2018-05-09 17:15 被阅读0次

    import java.util.concurrent.*;

    import java.util.concurrent.atomic.AtomicInteger;

    /**

    * Created by yehan .

    *  代码和思路主要来自于:

    *

    * tomcat :

    *        org.apache.catalina.core.StandardThreadExecutor

    *

    * java.util.concurrent

    * threadPoolExecutor execute执行策略:        优先offer到queue,queue满后再扩充线程到maxThread,如果已经到了maxThread就reject

    *                              比较适合于CPU密集型应用(比如runnable内部执行的操作都在JVM内部,memory copy, or compute等等)

    *

    * StandardThreadExecutor execute执行策略: 优先扩充线程到maxThread,再offer到queue,如果满了就reject

    *                          比较适合于业务处理需要远程资源的场景

    *

    * @author yehan

    *

    */

    public class StandardThreadExecutorextends ThreadPoolExecutor {

    public static final int DEFAULT_MIN_THREADS =20;

        public static final int DEFAULT_MAX_THREADS =200;

        /**1 minutes*/

        public static final int DEFAULT_MAX_IDLE_TIME =60 *1000;

        /**正在处理的任务数*/

        protected AtomicIntegersubmittedTasksCount;

        /**最大允许同时处理的任务数*/

        private int maxSubmittedTaskCount;

        public StandardThreadExecutor() {

    this(DEFAULT_MIN_THREADS, DEFAULT_MAX_THREADS);

        }

    public StandardThreadExecutor(int coreThread, int maxThreads) {

    this(coreThread, maxThreads, maxThreads);

        }

    public StandardThreadExecutor(int coreThread, int maxThreads, long keepAliveTime, TimeUnit unit) {

    this(coreThread, maxThreads, keepAliveTime, unit, maxThreads);

        }

    public StandardThreadExecutor(int coreThreads, int maxThreads, int queueCapacity) {

    this(coreThreads, maxThreads, queueCapacity, Executors.defaultThreadFactory());

        }

    public StandardThreadExecutor(int coreThreads, int maxThreads, int queueCapacity, ThreadFactory threadFactory) {

    this(coreThreads, maxThreads, DEFAULT_MAX_IDLE_TIME, TimeUnit.MILLISECONDS, queueCapacity, threadFactory);

        }

    public StandardThreadExecutor(int coreThreads, int maxThreads, long keepAliveTime, TimeUnit unit, int queueCapacity) {

    this(coreThreads, maxThreads, keepAliveTime, unit, queueCapacity, Executors.defaultThreadFactory());

        }

    public StandardThreadExecutor(int coreThreads, int maxThreads, long keepAliveTime, TimeUnit unit,

                                      int queueCapacity, ThreadFactory threadFactory) {

    this(coreThreads, maxThreads, keepAliveTime, unit, queueCapacity, threadFactory, new AbortPolicy());

        }

    public StandardThreadExecutor(int coreThreads, int maxThreads, long keepAliveTime, TimeUnit unit,

                                      int queueCapacity, ThreadFactory threadFactory, RejectedExecutionHandler handler) {

    super(coreThreads, maxThreads, keepAliveTime, unit, new ExecutorQueue(), threadFactory, handler);

            ((ExecutorQueue) getQueue()).setStandardThreadExecutor(this);

            submittedTasksCount =new AtomicInteger(0);

            // 最大并发任务限制: 队列buffer数 + 最大线程数

            maxSubmittedTaskCount = queueCapacity + maxThreads;

        }

    @Override

        public void execute(Runnable command) {

    int count =submittedTasksCount.incrementAndGet();

            // 超过最大的并发任务限制,进行 reject

    // 依赖的LinkedTransferQueue没有长度限制,因此这里进行控制

            if (count >maxSubmittedTaskCount) {

    submittedTasksCount.decrementAndGet();

                getRejectedExecutionHandler().rejectedExecution(command, this);

            }

    try {

    super.execute(command);

            }catch (RejectedExecutionException rx) {

    // there could have been contention around the queue

                if (!((ExecutorQueue) getQueue()).force(command)) {

    submittedTasksCount.decrementAndGet();

                    getRejectedExecutionHandler().rejectedExecution(command, this);

                }

    }

    }

    public int getSubmittedTasksCount() {

    return this.submittedTasksCount.get();

        }

    public int getMaxSubmittedTaskCount() {

    return maxSubmittedTaskCount;

        }

    @Override

        protected void afterExecute(Runnable r, Throwable t) {

    submittedTasksCount.decrementAndGet();

        }

    }

    /**

    * LinkedTransferQueue 能保证更高性能,相比与LinkedBlockingQueue有明显提升

    *

    *

    
    

    *        1) 不过LinkedTransferQueue的缺点是没有队列长度控制,需要在外层协助控制

    *

    *

    * @author yangnan

    *

    */

    class ExecutorQueueextends LinkedTransferQueue {

    private static final long serialVersionUID = -265236426751004839L;

        StandardThreadExecutorthreadPoolExecutor;

        public ExecutorQueue() {

    super();

        }

    public void setStandardThreadExecutor(StandardThreadExecutor threadPoolExecutor) {

    this.threadPoolExecutor = threadPoolExecutor;

        }

    /** 注:代码来源于 tomcat*/

        public boolean force(Runnable o) {

    if (threadPoolExecutor.isShutdown()) {

    throw new RejectedExecutionException("Executor not running, can't force a command into the queue");

            }

    // forces the item onto the queue, to be used if the task is rejected

            return super.offer(o);

        }

    /**

    *注:tomcat的代码进行一些小变更

    *

    */

        @Override

        public boolean offer(Runnable o) {

    int poolSize =threadPoolExecutor.getPoolSize();

            // we are maxed out on threads, simply queue the object

            if (poolSize ==threadPoolExecutor.getMaximumPoolSize()) {

    return super.offer(o);

            }

    // we have idle threads, just add it to the queue

    // note that we don't use getActiveCount(), see BZ 49730

            if (threadPoolExecutor.getSubmittedTasksCount() <= poolSize) {

    return super.offer(o);

            }

    // if we have less threads than maximum force creation of a new

    // thread

            if (poolSize

    return false;

            }

    // if we reached here, we need to add it to the queue

            return super.offer(o);

        }

    }

    相关文章

      网友评论

          本文标题:重写并优化 ThreadPoolExecutor 线程池

          本文链接:https://www.haomeiwen.com/subject/dalorftx.html