利用多线程进行多核计算
1、利用多线程进行多核计算，利用多核计算提升计算性能。
2、通过CountDownLatch countDownLatch = new CountDownLatch(threadSize);来将多个计算线程结果返回给主线程进行合并。
3、通过单例来生产固定大小线程池
private static Logger logger = LoggerFactory.getLogger(ThreadTool.class);
private static ThreadTool threadTool;
private ExecutorService executorService;
private ThreadTool(){
NamedThreadFactory threadFactory = new NamedThreadFactory("Gbdt-Thread-", true);
RejectedExecutionHandler handler = new RejectedExecutionHandler() {
private int i = 1;
@Override
public void rejectedExecution(Runnable r, ThreadPoolExecutor executor) {
if (i++ % 7 == 0) {
i = 1;
logger.warn("[Gbdt-23002]Task:{} has been reject for ThreadPool exhausted!" +
" pool:{}, active:{}, queue:{}, taskcnt: {}",
new Object[]{
r,
executor.getPoolSize(),
executor.getActiveCount(),
executor.getQueue().size(),
executor.getTaskCount()
});
}
throw new RejectedExecutionException("[Gbdt-23003]Gbdt thread pool of provider has bean exhausted");
}
};
executorService = new ThreadPoolExecutor(20, 500,
60, TimeUnit.MILLISECONDS,
new LinkedBlockingQueue<Runnable>(), threadFactory, handler);
}
public static ThreadTool getInstance(){
synchronized (ThreadTool.class) {
if (threadTool == null) { //Double Checked
threadTool = new ThreadTool();
}
}
return threadTool;
}
public ExecutorService getExecutorService(){
return executorService;
}4、获得线程池并将任务线程添加到线程中。
ExecutorService executorService= threadTool.getExecutorService();
executorService.execute(thread);
5、线程将执行完后将
// 倒数器减1
countDownLatch.countDown();6、countDownLatch设置等待时间countDownLatch减到0或超时则返回。
try
{
// 阻塞当前线程，直到倒数计数器倒数到0
countDownLatch.await(200, TimeUnit.MILLISECONDS);
}
catch (InterruptedException e)
{
log.error("线程计算异常:"+e.getMessage(),e);
}本文参与 腾讯云自媒体分享计划，分享自作者个人站点/博客。
原始发表：2017-07-12 ，如有侵权请联系 cloudcommunity@tencent.com 删除
评论
登录后参与评论
推荐阅读