解决InheritableThreadLocal线程池复用线程时，取值错误

一、前言

场景1：均是单线程情况

直接使用ThreadLocal及可以实现

场景2：存在异步情况

1. 仅兼容 new Thread()
2. 兼容 new Thread() 和 线程池

问题分析

1. 如果只使用原生的ThreadLocal，那么在异步的场景下ThreadLocal#get()方法则获取不到，所以采用InheritableThreadLocal代替ThreadLocal
2. InheritableThreadLocal，则可以解决兼容 new Thread()场景
3. InheritableThreadLocal不能解决线程场景，因为线程池存在线程复用问题，不会每次new，所以线程取InheritableThreadLocal 的值还是残留第一次传递进来值，导致结果不正确

问题复现

    private static final ThreadLocal<String> contextHolder = new InheritableThreadLocal<String>();
    public static void main(String[] args) {
        List<String> list = new ArrayList<>();
        for (int i = 1; i <= 10; i++) {
            list.add(i + "");
        }
        ThreadPoolTaskExecutor executor = build();
        list.forEach(id -> {
            contextHolder.set(id);
            System.out.println(Thread.currentThread().getName() + "----" + contextHolder.get());
            executor.execute(() -> System.out.println(Thread.currentThread().getName() + "----" + contextHolder.get()));
        });
    }
    private static ThreadPoolTaskExecutor build() {
        ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
        // 核心线程池大小
        executor.setCorePoolSize(5);
        // 最大可创建的线程数
        executor.setMaxPoolSize(10);
        // 队列最大长度
        executor.setQueueCapacity(1000);
        // 线程池维护线程所允许的空闲时间
        executor.setKeepAliveSeconds(60);
        executor.setThreadNamePrefix("taskExecutor-");
        executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
        executor.setWaitForTasksToCompleteOnShutdown(true);
        executor.setAwaitTerminationSeconds(60);
        executor.initialize();
        return executor;
    }

输出结果：

main----1
main----2
main----3
taskExecutor-1----1
main----4
taskExecutor-2----2
taskExecutor-3----3
main----5
taskExecutor-4----4
main----6
main----7
taskExecutor-1----1
taskExecutor-2----2
main----8
taskExecutor-5----5
taskExecutor-3----3
main----9
main----10
taskExecutor-4----4
taskExecutor-1----1

二、解决方案

1. maven依赖

	...
    <dependencies>
        <dependency>
            <groupId>org.springframework</groupId>
            <artifactId>spring-context</artifactId>
            <version>5.1.7.RELEASE</version>
        </dependency>
        <dependency>
            <groupId>com.alibaba</groupId>
            <artifactId>transmittable-thread-local</artifactId>
            <version>2.2.0</version>
        </dependency>
    </dependencies>

2. 解决方案——源码

import com.alibaba.ttl.TransmittableThreadLocal;
import com.alibaba.ttl.threadpool.TtlExecutors;
import org.springframework.scheduling.concurrent.ThreadPoolTaskExecutor;

import java.util.ArrayList;
import java.util.List;
import java.util.concurrent.Executor;
import java.util.concurrent.ThreadPoolExecutor;

public class TestDemo {
    private static final ThreadLocal<String> contextHolder = new TransmittableThreadLocal();

    public static void main(String[] args) {
        List<String> list = new ArrayList<>();
        for (int i = 1; i <= 10; i++) {
            list.add(i + "");
        }
        Executor executor = build();
        list.forEach(id -> {
            contextHolder.set(id);
            System.out.println(Thread.currentThread().getName() + "----" + contextHolder.get());
            Runnable runnable = () -> {
                System.out.println(Thread.currentThread().getName() + "----" + contextHolder.get());
            };
            executor.execute(runnable);
            new Thread(runnable).start();
        });
    }

    private static Executor build() {
        ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
        // 核心线程池大小
        executor.setCorePoolSize(5);
        // 最大可创建的线程数
        executor.setMaxPoolSize(10);
        // 队列最大长度
        executor.setQueueCapacity(1000);
        // 线程池维护线程所允许的空闲时间
        executor.setKeepAliveSeconds(60);
        executor.setThreadNamePrefix("taskExecutor-");
        executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
        executor.setWaitForTasksToCompleteOnShutdown(true);
        executor.setAwaitTerminationSeconds(60);
        executor.initialize();
        return TtlExecutors.getTtlExecutor(executor);
    }
}

3. 输出结果

main----1
main----2
Thread-1----1
taskExecutor-1----1
main----3
Thread-2----2
taskExecutor-2----2
main----4
taskExecutor-3----3
Thread-3----3
main----5
taskExecutor-4----4
Thread-4----4
main----6
taskExecutor-5----5
Thread-5----5
taskExecutor-1----6
main----7
Thread-6----6
taskExecutor-2----7
main----8
Thread-7----7
taskExecutor-3----8
main----9
Thread-8----8
taskExecutor-4----9
main----10
Thread-9----9
taskExecutor-5----10
Thread-10----10