批量操作

private void batchDeal(List data, int batchNum) throws InterruptedException {
int totalNum = data.size();
int pageNum = totalNum % batchNum == 0 ? totalNum / batchNum : totalNum / batchNum + 1;
ExecutorService executor = Executors.newFixedThreadPool(pageNum);
try {
CountDownLatch countDownLatch = new CountDownLatch(pageNum);
List subData = null;
int fromIndex, toIndex;
for (int i = 0; i < pageNum; i++) {
fromIndex = i * batchNum;
toIndex = Math.min(totalNum, fromIndex + batchNum);
subData = data.subList(fromIndex, toIndex);
ImportTask task = new ImportTask(subData, countDownLatch);
executor.execute(task);
}
// 主线程必须在启动其它线程后立即调用CountDownLatch.await()方法，
// 这样主线程的操作就会在这个方法上阻塞，直到其它线程完成各自的任务。
// 计数器的值等于0时，主线程就能通过await()方法恢复执行自己的任务。
countDownLatch.await();
logger.info("数据操作完成!可以在此开始其它业务");
} finally {
// 关闭线程池，释放资源
executor.shutdown();
}
}

class ImportTask implements Runnable {
private List list;
private CountDownLatch countDownLatch;

public ImportTask(List data, CountDownLatch countDownLatch) {
this.list = data;
this.countDownLatch = countDownLatch;
}

@Override
public void run() {
if (null != list) {
// 业务逻辑，例如批量insert或者update
logger.info("现在操作的数据是{}", list);
}
// 发出线程任务完成的信号
countDownLatch.countDown();
}
}

转载于:https://www.cnblogs.com/yuruyi/p/11447603.html