针对上面的场景我们可以通过 KeyAffinityExecutor (KeyAffinityExecutor 是一个可以按照指定的Key亲和顺序消费的执行器) 来解决这个问题，我们下面一起来了解下 KeyAffinityExecutor 。
基本使用
导入依赖
com.github.phantomthief
more-lambdas
0.1.55
创建线程池
public class KeyAffinityExecutorTest {
@Test
public void submitTaskKeyAffinityExecutor() {
//线程池
KeyAffinityExecutor keyAffinityExecutor = KeyAffinityExecutor
.newSerializingExecutor(2, 200, "测试-%d");
//需要下单的信息
List orders = new ArrayList<>();
orders.add(new Order(1, "iPhone 16 Max"));
orders.add(new Order(1, "Thinking In Java"));
orders.add(new Order(1, "MengNiu Milk"));
orders.add(new Order(2, "Thinking In Java"));
orders.add(new Order(3, "HUAWEI 100P"));
orders.add(new Order(4, "XIAOMI 20"));
orders.add(new Order(5, "OPPO 98"));
orders.add(new Order(6, "HP EC80"));
orders.add(new Order(7, "BBK 100P"));
orders.add(new Order(8, "TCL 1380"));
orders.add(new Order(9, "CHANGHONG 32"));
orders.forEach(order -> keyAffinityExecutor.submit(order.getAccountId(), () -> {
System.out.println(Thread.currentThread() + " accountId:" + order.getAccountId() +
", skuNo:" + order.getSkuNo() + " checkout success!");
return null;
}));
try {
Thread.sleep(1000L);
} catch (InterruptedException e) {
throw new RuntimeException(e);
}
Assert.assertTrue(true);
}
@Data
@AllArgsConstructor
public static class Order {
long accountId;
String skuNo;
}
} 输出结果如下：
Thread[测试-0,5,main] accountId:1, skuNo:iPhone 16 Max checkout success!
Thread[测试-1,5,main] accountId:2, skuNo:Thinking In Java checkout success!
Thread[测试-1,5,main] accountId:3, skuNo:HUAWEI 100P checkout success!
Thread[测试-1,5,main] accountId:4, skuNo:XIAOMI 20 checkout success!
Thread[测试-0,5,main] accountId:1, skuNo:Thinking In Java checkout success!
Thread[测试-1,5,main] accountId:6, skuNo:HP EC80 checkout success!
Thread[测试-0,5,main] accountId:1, skuNo:MengNiu Milk checkout success!
Thread[测试-1,5,main] accountId:8, skuNo:TCL 1380 checkout success!
Thread[测试-0,5,main] accountId:5, skuNo:OPPO 98 checkout success!
Thread[测试-0,5,main] accountId:7, skuNo:BBK 100P checkout success!
Thread[测试-0,5,main] accountId:9, skuNo:CHANGHONG 32 checkout success!结论：对于 acccountId = 1 有三条数据都是在同一个线程下面执行，线程ID：测试-0 所以可以保证局部有序。
实现原理
- 选择执行的线程池， 这里我们可以看到，如果当前 key 存在线程池就直接返回，如果不存在就创建，或者选择一个任务比较少的线程池，这里可以保证任务分发的均匀性。
//通过 key 选出一个执行线程
@Nonnull
public V select(K key) {
int thisCount = count.getAsInt();
tryCheckCount(thisCount);
KeyRef keyRef = mapping.compute(key, (k, v) -> {
// 如果不存在就创建一个
if (v == null) {
if (usingRandom.test(thisCount)) {
do {
try {
v = new KeyRef(all.get(ThreadLocalRandom.current().nextInt(all.size())));
} catch (IndexOutOfBoundsException e) {
// ignore
}
} while (v == null);
} else {
v = all.stream()
.min(comparingInt(ValueRef::concurrency))
.map(KeyRef::new)
.orElseThrow(IllegalStateException::new);
}
}
v.incrConcurrency();
return v;
});
return keyRef.ref();
}- 执行线程池的初始化, 这里的本质是创建只有一个线程的线程池。这样就可以保证，任务被路由到同一个 key 下面，那么就可以保证顺序执行。
static Supplier executor(String threadName, int queueBufferSize) {
return new Supplier() {
// ThreadFactory
private final ThreadFactory threadFactory = new ThreadFactoryBuilder()
.setNameFormat(threadName)
.build();
@Override
public ExecutorService get() {
LinkedBlockingQueue queue;
if (queueBufferSize > 0) {
// blockingQueue
queue = new LinkedBlockingQueue(queueBufferSize) {
@Override
public boolean offer(Runnable e) {
try {
//让 offer 方法阻塞，
//为什么这么做可以看 ThreadPoolExecutor 1347 行
put(e);
return true;
} catch (InterruptedException ie) {
Thread.currentThread().interrupt();
}
return false;
}
};
} else {
queue = new LinkedBlockingQueue<>();
}
//创建一个线程的线程池
return new ThreadPoolExecutor(1, 1, 0L, MILLISECONDS, queue, threadFactory);
}
};
} - 最后任务执行完毕，回收线程。
//当每一个key执行完之后回收处理这个key的线程池.
public void finishCall(K key) {
//如果执行完毕后返回 null
mapping.computeIfPresent(key, (k, v) -> {
if (v.decrConcurrency()) {
return null;
} else {
return v;
}
});
}总结，这里其实我们也可以通过只有一个线程的线程数组实现，来实现按照唯一key，进行 hash 路由。
