面试题：Java可缓存线程池资源管理的性能优化

线程创建与销毁的优化

重用线程：可缓存线程池默认会在一定时间（60秒）内缓存空闲线程。为了减少线程创建与销毁开销，确保业务场景下线程的复用率足够高。
- 代码实现：使用Executors.newCachedThreadPool()创建线程池时，系统会自动管理线程的缓存与复用。
```
ExecutorService executorService = Executors.newCachedThreadPool();
for (int i = 0; i < 10; i++) {
    executorService.submit(() -> {
        // 执行任务
        System.out.println(Thread.currentThread().getName() + " is running task");
    });
}
executorService.shutdown();
```

限制线程创建数量：虽然可缓存线程池理论上可创建无限线程，但在高并发场景可能导致资源耗尽。可以自定义线程池并设置最大线程数。

代码实现：

ThreadPoolExecutor executor = new ThreadPoolExecutor(
    0, // 核心线程数为0
    100, // 最大线程数为100
    60L, TimeUnit.SECONDS,
    new SynchronousQueue<>());
for (int i = 0; i < 10; i++) {
    executor.submit(() -> {
        // 执行任务
        System.out.println(Thread.currentThread().getName() + " is running task");
    });
}
executor.shutdown();

任务调度优化

优先队列调度：对于有优先级的任务，可以使用PriorityBlockingQueue作为任务队列，让线程池按照任务优先级执行。

代码实现：

class PriorityTask implements Runnable, Comparable<PriorityTask> {
    private int priority;
    private String taskName;
    public PriorityTask(int priority, String taskName) {
        this.priority = priority;
        this.taskName = taskName;
    }
    @Override
    public void run() {
        System.out.println(taskName + " with priority " + priority + " is running");
    }
    @Override
    public int compareTo(PriorityTask other) {
        return Integer.compare(this.priority, other.priority);
    }
}
PriorityBlockingQueue<Runnable> taskQueue = new PriorityBlockingQueue<>();
ThreadPoolExecutor executor = new ThreadPoolExecutor(
    0, 100, 60L, TimeUnit.SECONDS,
    taskQueue);
executor.submit(new PriorityTask(3, "Task 3"));
executor.submit(new PriorityTask(1, "Task 1"));
executor.submit(new PriorityTask(2, "Task 2"));
executor.shutdown();

分队列调度：根据任务类型将任务分配到不同队列，每个队列可配置不同的线程数处理。

代码实现：

class Type1Task implements Runnable {
    @Override
    public void run() {
        System.out.println("Type 1 task is running");
    }
}
class Type2Task implements Runnable {
    @Override
    public void run() {
        System.out.println("Type 2 task is running");
    }
}
BlockingQueue<Runnable> type1Queue = new ArrayBlockingQueue<>(100);
BlockingQueue<Runnable> type2Queue = new ArrayBlockingQueue<>(100);
ThreadPoolExecutor type1Executor = new ThreadPoolExecutor(
    5, 10, 60L, TimeUnit.SECONDS, type1Queue);
ThreadPoolExecutor type2Executor = new ThreadPoolExecutor(
    3, 5, 60L, TimeUnit.SECONDS, type2Queue);
type1Executor.submit(new Type1Task());
type2Executor.submit(new Type2Task());
type1Executor.shutdown();
type2Executor.shutdown();

其他优化策略

线程池监控与调整：使用ThreadPoolExecutor提供的监控方法，如getActiveCount()、getTaskCount()等，实时了解线程池运行状态，动态调整线程池参数。

代码实现：

ThreadPoolExecutor executor = new ThreadPoolExecutor(
    0, 100, 60L, TimeUnit.SECONDS, new SynchronousQueue<>());
for (int i = 0; i < 10; i++) {
    executor.submit(() -> {
        // 执行任务
        System.out.println(Thread.currentThread().getName() + " is running task");
    });
}
System.out.println("Active threads: " + executor.getActiveCount());
System.out.println("Total tasks: " + executor.getTaskCount());
executor.shutdown();

合理设置线程存活时间：可缓存线程池默认线程存活时间为60秒，可根据业务负载情况适当调整这个时间。如果任务执行频率高且持续时间短，可适当延长存活时间；如果任务执行频率低，可缩短存活时间以释放资源。
- 代码实现：
```
ThreadPoolExecutor executor = new ThreadPoolExecutor(
    0, 100, 30L, TimeUnit.SECONDS, new SynchronousQueue<>());
// 执行任务
executor.shutdown();
```

星途面试题库

面试题：Java可缓存线程池资源管理的性能优化

知识考点

面试题答案

线程创建与销毁的优化

任务调度优化

其他优化策略