Eureka中客户端缓存刷新、心跳监测,调远程服务端接口操作的定时任务,调远程网络不确定性导致定时任务执行不确定性,Eureka设计了一个自动调节间隔的周期性任务!!!通过一个线程类的run方法的finally中再次调用该线程实现。
DiscoveryClient类里的initScheduledTasks()方法:该方法有两处使用了该定时任务线程类。
private void initScheduledTasks() {
if (clientConfig.shouldFetchRegistry()) {
// registry cache refresh timer
int registryFetchIntervalSeconds = clientConfig.getRegistryFetchIntervalSeconds();
int expBackOffBound = clientConfig.getCacheRefreshExecutorExponentialBackOffBound();
scheduler.schedule(
new TimedSupervisorTask(
"cacheRefresh",
scheduler,
cacheRefreshExecutor,
registryFetchIntervalSeconds,
TimeUnit.SECONDS,
expBackOffBound,
new CacheRefreshThread()//该线程执行更新的具体逻辑
),
registryFetchIntervalSeconds, TimeUnit.SECONDS);
}
if (clientConfig.shouldRegisterWithEureka()) {
int renewalIntervalInSecs = instanceInfo.getLeaseInfo().getRenewalIntervalInSecs();
int expBackOffBound = clientConfig.getHeartbeatExecutorExponentialBackOffBound();
logger.info("Starting heartbeat executor: " + "renew interval is: {}", renewalIntervalInSecs);
// Heartbeat timer
scheduler.schedule(
new TimedSupervisorTask(
"heartbeat",
scheduler,
heartbeatExecutor,
renewalIntervalInSecs,
TimeUnit.SECONDS,
expBackOffBound,
new HeartbeatThread()//该线程执行续约的具体逻辑
),
renewalIntervalInSecs, TimeUnit.SECONDS);
…………省略
}
上述代码中,scheduler是ScheduledExecutorService接口的实现:因为两处使用,所以创建了2个线程的线程池。且在new子线程的时候将scheduler传入。
scheduler = Executors.newScheduledThreadPool(2,
new ThreadFactoryBuilder()
.setNameFormat("DiscoveryClient-%d")
.setDaemon(true)
.build());
其schedule方法的官方文档如下所示:
创建了一个一次性动作,那是怎么实现循环定时执行的?答案就在TimedSupervisorTask线程类的run方法里》
亮点就在run方法的finally里的
scheduler.schedule(this, delay.get(), TimeUnit.MILLISECONDS);
package com.netflix.discovery;
import java.util.TimerTask;
import java.util.concurrent.Future;
import java.util.concurrent.RejectedExecutionException;
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.ThreadPoolExecutor;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;
import java.util.concurrent.atomic.AtomicLong;
import com.netflix.servo.monitor.Counter;
import com.netflix.servo.monitor.LongGauge;
import com.netflix.servo.monitor.MonitorConfig;
import com.netflix.servo.monitor.Monitors;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
/**
* A supervisor task that schedules subtasks while enforce a timeout.
* Wrapped subtasks must be thread safe.
*
* @author David Qiang Liu
*/
public class TimedSupervisorTask extends TimerTask {
private static final Logger logger = LoggerFactory.getLogger(TimedSupervisorTask.class);
private final Counter timeoutCounter;
private final Counter rejectedCounter;
private final Counter throwableCounter;
private final LongGauge threadPoolLevelGauge;
private final ScheduledExecutorService scheduler;
private final ThreadPoolExecutor executor;
private final long timeoutMillis;
private final Runnable task;
private final AtomicLong delay;
private final long maxDelay;
public TimedSupervisorTask(String name, ScheduledExecutorService scheduler, ThreadPoolExecutor executor,
int timeout, TimeUnit timeUnit, int expBackOffBound, Runnable task) {
this.scheduler = scheduler;
this.executor = executor;
this.timeoutMillis = timeUnit.toMillis(timeout);
this.task = task;
this.delay = new AtomicLong(timeoutMillis);
this.maxDelay = timeoutMillis * expBackOffBound;
// Initialize the counters and register.
timeoutCounter = Monitors.newCounter("timeouts");
rejectedCounter = Monitors.newCounter("rejectedExecutions");
throwableCounter = Monitors.newCounter("throwables");
threadPoolLevelGauge = new LongGauge(MonitorConfig.builder("threadPoolUsed").build());
Monitors.registerObject(name, this);
}
@Override
public void run() {
Future> future = null;
try {
//使用Future,可以设定子线程的超时时间,这样当前线程就不用无限等待了
future = executor.submit(task);
threadPoolLevelGauge.set((long) executor.getActiveCount());
//指定等待子线程的最长时间
future.get(timeoutMillis, TimeUnit.MILLISECONDS); // block until done or timeout
//delay是个关键变量,后面会用到,
//这里每次执行任务成功都会将delay重置,即时间间隔还原到最初值。
delay.set(timeoutMillis);
threadPoolLevelGauge.set((long) executor.getActiveCount());
} catch (TimeoutException e) {
logger.warn("task supervisor timed out", e);
timeoutCounter.increment();
long currentDelay = delay.get();
//任务线程超时的时候,就把delay变量翻倍,但不会超过外部调用时设定的最大延时时间
long newDelay = Math.min(maxDelay, currentDelay * 2);
//设置为最新的值,考虑到多线程,所以用了CAS
delay.compareAndSet(currentDelay, newDelay);
} catch (RejectedExecutionException e) {
if (executor.isShutdown() || scheduler.isShutdown()) {
logger.warn("task supervisor shutting down, reject the task", e);
} else {
logger.warn("task supervisor rejected the task", e);
}
rejectedCounter.increment();
} catch (Throwable e) {
//一旦线程池的阻塞队列中放满了待处理任务,触发了拒绝策略,就会将调度器停掉
if (executor.isShutdown() || scheduler.isShutdown()) {
logger.warn("task supervisor shutting down, can't accept the task");
} else {
logger.warn("task supervisor threw an exception", e);
}
throwableCounter.increment();
} finally {
//这里任务要么执行完毕,要么发生异常,都用cancel方法来清理任务;
if (future != null) {
future.cancel(true);
}
//只要调度器没有停止,就在指定等待时间之后再执行一次同样的任务
if (!scheduler.isShutdown()) {
//假设外部调用时传入的超时时间为30秒(构造方法的入参timeout),最大间隔时间为50秒(构造方法的入参expBac kOffBound)
//如果最近一次任务没有超时,那么就在30秒后开始新任务,
//如果最近一次任务超时了,那么就在50秒后开始新任务(异常处理中有个乘以二的操作,乘以二后的60秒超过了最大 间隔50秒)
// TODO 亮点所在!!!再次调用。
scheduler.schedule(this, delay.get(), TimeUnit.MILLISECONDS);
}
}
}
}
TimerTask是Java的util包中Runnable的实现类:
public abstract class TimerTask implements Runnable {
真相就在上面的最后一行代码中:scheduler.schedule(this, delay.get(), TimeUnit.MILLISECONDS):执行完任务后,会再次调用schedule方法,在指定的时间之后执行一次相同的任务,这个间隔时间和最近一次任务是否超时有关,如果超时了间隔时间就会变大。