springboot中@Async默认线程池导致OOM问题

前言：

1.最近项目上在测试人员压测过程中发现了OOM问题，项目使用springboot搭建项目工程，通过查看日志中包含信息：unable to create new native thread

内存溢出的三种类型：
1.第一种OutOfMemoryError： PermGen space，发生这种问题的原意是程序中使用了大量的jar或class
2.第二种OutOfMemoryError： Java heap space，发生这种问题的原因是java虚拟机创建的对象太多
3.第三种OutOfMemoryError：unable to create new native thread，创建线程数量太多，占用内存过大

初步分析：

1.初步怀疑是线程创建太多导致，使用jstack 线程号 > /tmp/oom.log将应用的线程信息打印出来。查看oom.log，发现大量线程处于Runnable状态，基本可以确认是线程创建太多了。

代码分析：

1.出问题的微服务是日志写库服务，对比日志，锁定在writeLog方法上，wirteLog方法使用spring-@Async注解，写库操作采用的是异步写入方式。
2.之前没有对@Async注解深入研究过，只是知道可以自定义内部线程池，经查看，日志写库服务并未自定义异步配置，使用的是spring-@Async默认异步配置
3.首先简单百度了下，网上提到@Async默认异步配置使用的是SimpleAsyncTaskExecutor，该线程池默认来一个任务创建一个线程，在压测情况下，会有大量写库请求进入日志写库服务，这时就会不断创建大量线程，极有可能压爆服务器内存。

借此机会也学习了下SimpleAsyncTaskExecutor源码，总结如下：

1.SimpleAsyncTaskExecutor提供了限流机制，通过concurrencyLimit属性来控制开关，当concurrencyLimit>=0时开启限流机制，默认关闭限流机制即concurrencyLimit=-1，当关闭情况下，会不断创建新的线程来处理任务，核心代码如下：

public void execute(Runnable task, long startTimeout) {
 Assert.notNull(task, "Runnable must not be null");
 Runnable taskToUse = (this.taskDecorator != null ? this.taskDecorator.decorate(task) : task);
 //判断是否开启限流机制
 if (isThrottleActive() && startTimeout > TIMEOUT_IMMEDIATE) {
  //执行前置操作，进行限流
  this.concurrencyThrottle.beforeAccess();
  //执行完线程任务，会执行后置操作concurrencyThrottle.afterAccess()，配合进行限流
  doExecute(new ConcurrencyThrottlingRunnable(taskToUse));
 }
 else {
  doExecute(taskToUse);
 }
}

2.SimpleAsyncTaskExecutor限流实现

首先任务进来，会循环判断当前执行线程数是否超过concurrencyLimit，如果超了，则当前线程调用wait方法，释放monitor对象锁，进入等待

protected void beforeAccess() {
if (this.concurrencyLimit == NO_CONCURRENCY) {
throw new IllegalStateException(
"Currently no invocations allowed - concurrency limit set to NO_CONCURRENCY");
}
if (this.concurrencyLimit > 0) {
boolean debug = logger.isDebugEnabled();
synchronized (this.monitor) {
boolean interrupted = false;
while (this.concurrencyCount >= this.concurrencyLimit) {
if (interrupted) {
throw new IllegalStateException("Thread was interrupted while waiting for invocation access, " +
"but concurrency limit still does not allow for entering");
}
if (debug) {
logger.debug("Concurrency count " + this.concurrencyCount +
" has reached limit " + this.concurrencyLimit + " - blocking");
}
try {
this.monitor.wait();
}
catch (InterruptedException ex) {
// Re-interrupt current thread, to allow other threads to react.
Thread.currentThread().interrupt();
interrupted = true;
}
}
if (debug) {
logger.debug("Entering throttle at concurrency count " + this.concurrencyCount);
}
this.concurrencyCount++;
}
}
}

2.SimpleAsyncTaskExecutor限流实现：首先任务进来，会循环判断当前执行线程数是否超过concurrencyLimit，如果超了，则当前线程调用wait方法，释放monitor对象锁，进入等待状态。

protected void beforeAccess() {
if (this.concurrencyLimit == NO_CONCURRENCY) {
throw new IllegalStateException(
"Currently no invocations allowed - concurrency limit set to NO_CONCURRENCY");
}
if (this.concurrencyLimit > 0) {
boolean debug = logger.isDebugEnabled();
synchronized (this.monitor) {
boolean interrupted = false;
while (this.concurrencyCount >= this.concurrencyLimit) {
if (interrupted) {
throw new IllegalStateException("Thread was interrupted while waiting for invocation access, " +
"but concurrency limit still does not allow for entering");
}
if (debug) {
logger.debug("Concurrency count " + this.concurrencyCount +
" has reached limit " + this.concurrencyLimit + " - blocking");
}
try {
this.monitor.wait();
}
catch (InterruptedException ex) {
// Re-interrupt current thread, to allow other threads to react.
Thread.currentThread().interrupt();
interrupted = true;
}
}
if (debug) {
logger.debug("Entering throttle at concurrency count " + this.concurrencyCount);
}
this.concurrencyCount++;
}
}
}

线程任务执行完毕后，当前执行线程数会减一，会调用monitor对象的notify方法，唤醒等待状态下的线程，等待状态下的线程会竞争monitor锁，竞争到，会继续执行线程任务。

protected void afterAccess() {
if (this.concurrencyLimit >= 0) {
synchronized (this.monitor) {
this.concurrencyCount--;
if (logger.isDebugEnabled()) {
logger.debug("Returning from throttle at concurrency count " + this.concurrencyCount);
}
this.monitor.notify();
}
}
}

虽然看了源码了解了SimpleAsyncTaskExecutor有限流机制，实践出真知，我们还是测试下：
一、测试未开启限流机制下，我们启动20个线程去调用异步方法，查看Java VisualVM工具如下：

二、测试开启限流机制，开启限流机制的代码如下:

@Configuration
@EnableAsync
public class AsyncCommonConfig extends AsyncConfigurerSupport {
 @Override
 public Executor getAsyncExecutor() {
   SimpleAsyncTaskExecutor executor = new SimpleAsyncTaskExecutor();
   //设置允许同时执行的线程数为10
executor.setConcurrencyLimit(10);
   return executor;
 }
}

同样，我们启动20个线程去调用异步方法，查看Java VisualVM工具如下：

通过上面验证可知：
1.开启限流情况下，能有效控制应用线程数
2.虽然可以有效控制线程数，但执行效率会降低，会出现主线程等待，线程竞争的情况。
3.限流机制适用于任务处理比较快的场景，对于应用处理时间比较慢的场景并不适用。==

最终解决办法：
1.自定义线程池，使用LinkedBlockingQueue阻塞队列来限定线程池的上限
2.定义拒绝策略，如果队列满了，则拒绝处理该任务，打印日志，代码如下：

public class AsyncConfig implements AsyncConfigurer{
 private Logger logger = LogManager.getLogger();

@Value("${thread.pool.corePoolSize:10}")
 private int corePoolSize;

@Value("${thread.pool.maxPoolSize:20}")
 private int maxPoolSize;

@Value("${thread.pool.keepAliveSeconds:4}")
 private int keepAliveSeconds;

@Value("${thread.pool.queueCapacity:512}")
 private int queueCapacity;

@Override
 public Executor getAsyncExecutor() {
   ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
   executor.setCorePoolSize(corePoolSize);
   executor.setMaxPoolSize(maxPoolSize);
   executor.setKeepAliveSeconds(keepAliveSeconds);
   executor.setQueueCapacity(queueCapacity);
   executor.setRejectedExecutionHandler((Runnable r, ThreadPoolExecutor exe) -> {
       logger.warn("当前任务线程池队列已满.");
   });
   executor.initialize();
   return executor;
 }

@Override
 public AsyncUncaughtExceptionHandler getAsyncUncaughtExceptionHandler() {
   return new AsyncUncaughtExceptionHandler() {
     @Override
     public void handleUncaughtException(Throwable ex , Method method , Object... params) {
       logger.error("线程池执行任务发生未知异常.", ex);
     }
   };
 }
}

来源：https://blog.csdn.net/ignorewho/article/details/85603920