java8 CompletableFuture.supplyAsync +線程池 實現多線程處理

public void dealGovernanceStrategyNew(List<StrategyStreamOperation> commonAll, StrategyDetail strategyDetail, List<String> instanceList) {
	if (Objects.isNull(strategyDetail.getType()) && Objects.isNull(strategyDetail.getRetainNum())) {
		// 500一組分批處理
		List<List<String>> lists = ListSplitUtil.splitList(instanceList, 500);
		// 對於集合寫操作：synchronizedList 相對於Vector 、CopyOnWriteArrayList性能更佳。 讀操作建議CopyOnWriteArrayList
		// 多線程批量處理數據（500一組） 這里使用java8 CompletableFuture.supplyAsync實現多線程處理
		List<StrategyStreamOperation> operationList = Collections.synchronizedList(new ArrayList<>());
		// 手動創建線程池
		ExecutorService executorService =  new ThreadPoolExecutor(lists.size(), lists.size(), 0, TimeUnit.MILLISECONDS,
				new LinkedBlockingQueue<Runnable>(1024), new ThreadPoolExecutor.AbortPolicy());
		lists.stream().forEach((List<String> item) ->{
			CompletableFuture<List<StrategyStreamOperation>> cf = CompletableFuture.supplyAsync(()->{
				// 持續天數的峰值判斷的
				List<StrategyStreamOperation> futureList = getStrategyByCommon();
				return futureList;
			}, executorService);
			try {
				// 獲取線程執行結果
				operationList.addAll(cf.get());
			} catch (InterruptedException e) {
				log.error("從線程中獲取執行結果失敗", e);
			} catch (ExecutionException e) {
				log.error("從線程中獲取執行結果失敗", e);
			}
		});
		// 關閉線程池,不再接收新任務
		executorService.shutdown();
		commonAll.addAll(operationList);
	}
}