同步锁:synchronized、Lock区别
1、synchronized是不需要进行手动解锁
2、synchronized可以锁方法、锁同步代码块
3、synchronized是Java自带关键字
4、Lock锁是一个类且它拥有synchronized的所有功能还具备扩展
5、Lock锁的实现类ReentrantLock可以实现公平和非公平锁
6、Lock锁需要手动加锁和手动解锁
7、synchronized不可中断而Lock锁可以实现中断
synchronized
Lock : 主要使用到的实现类ReentrantLock(可重入锁)
集合的线程不安全情况和解决方案
List : ArrayList不安全List,但是在单线程情况下是高效的!
多线程下错误案例:
List<Integer> list = new ArrayList<>();for (int i = 0; i < 30; i++) { final Integer temp = i; new Thread(()->{ list.add(temp); System.out.println(list); }, String.valueOf(temp)).start(); }//结果出现并发修改异常ConcurrentModificationException[0, 1, 2, 3, 5, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 19, 18, 20, 21, 23, 22, 24, 25, 26, 27, 29, 28]//Exception in thread "11" Exception in thread "15" Exception in thread "19" java.util.ConcurrentModificationException解决方案:
//1.使用集合安全类进行转换List<Integer> list = Collections.synchronizedList(new ArrayList<>()); for (int i = 0; i < 30; i++) { final Integer temp = i; new Thread(()->{ list.add(temp); System.out.println(list); }, String.valueOf(temp)).start(); }//2.使用List对应的Vector List<Integer> list = new Vector<>(); for (int i = 0; i < 30; i++) { final Integer temp = i; new Thread(()->{ list.add(temp); System.out.println(list); }, String.valueOf(temp)).start(); }//3.使用CopyOnWriteArrayListList<Integer> list = new CopyOnWriteArrayList<>(); for (int i = 0; i < 30; i++) { final Integer temp = i; new Thread(()->{ list.add(temp); System.out.println(list); }, String.valueOf(temp)).start(); }Set集合:HashSet
多线程下错误案例
Set<Integer> set = new HashSet<>(); for (int i = 0; i < 30; i++) { final Integer temp = i; new Thread(()->{ set.add(temp); System.out.println(set); }, String.valueOf(temp)).start(); }// 结果:抛出ConcurrentModificationException并发修改异常!解决方案:
// 1.使用Collections.synchronizedSet安全集合包装 Set<Integer> set = Collections.synchronizedSet(new HashSet<>());// 2.使用CopyOnWriteArraySetSet<Integer> set = new CopyOnWriteArraySet<>();map集合:HashMap
Map<String, Object> map = new HashMap<>(); for (int i = 0; i < 30; i++) { final Integer temp = i; new Thread(()->{ map.put(String.valueOf(temp), temp); System.out.println(map); }, String.valueOf(temp)).start(); }//结果:Exception in thread "6" java.util.ConcurrentModificationException并发修改异常解决方案:
// 1.ConcurrentHashMapMap<String, Object> map = new ConcurrentHashMap<>();for (int i = 0; i < 70; i++) { final Integer temp = i; new Thread(()->{ map.put(String.valueOf(temp), temp); System.out.println(map); }, String.valueOf(temp)).start();}// 2.Hashtable(效率低)CountDownLatch(减法计数器)
CountDownLatch countDownLatch = new CountDownLatch(10); // 传入一个数字,要执行多少次countDownLatch.countDown(); // 每次执行完一个任务后,进行减1操作countDownLatch.await(); // 等待计数器归零,只有等上面执行次数完毕后,才能执行后面的操作CyclicBarrier(加法计数器)
CyclicBarrier cyclicBarrier = new CyclicBarrier(10); // 初始化计数器容量,默认构造Runnable为null// 当计数器到达10的时候,就执行Runnable里面的具体操作CyclicBarrier cyclicBarrier = new CyclicBarrier(10, new Runnable() { @Override public void run() { System.out.println("executor other thing!"); } });cyclicBarrier.await(); // 等待计数器到达初始化计数器值,然后才能执行下面操作!栗子:
//初始化cyclicBarrier加法计数器CyclicBarrier cyclicBarrier = new CyclicBarrier(10, new Runnable() { @Override public void run() { System.out.println("执行到第十个啦,完结!"); } }); for (int i = 1; i <= 10; i++) { int u = i; new Thread(()->{ try { System.out.println("执行到第" + u +"个了, 还剩" + (10 - u) + "个"); //每执行完一个线程就进行加一操作!当执行完第十个就触发cyclicBarrier初始化中的Runnable接口实现 cyclicBarrier.await(); } catch (InterruptedException e) { e.printStackTrace(); } catch (BrokenBarrierException e) { e.printStackTrace(); } }).start(); }Semaphore(信号量)
Semaphore : 一般用于限流情况
semaphore.acquire():获得线程使用权限
semaphore.release():释放线程使用权限
Semaphore semaphore = new Semaphore(2); for (int i = 1; i <= 4; i++) { new Thread(() -> { try { // 得到线程执行权限,当线程数到达了信号量初始化容量,其他线程就会等待(阻塞)当前线程执行完毕并释放执行权限才可继续执行! semaphore.acquire(); System.out.println("当前线程:" + Thread.currentThread().getName() + "开始执行..."); TimeUnit.SECONDS.sleep(2); System.out.println("当前线程:" + Thread.currentThread().getName() + "执行完毕"); } catch (InterruptedException e) { e.printStackTrace(); } finally { semaphore.release(); //释放线程执行权限 } }, String.valueOf(i)).start(); } }// 结果当前线程:1开始执行...当前线程:2开始执行...当前线程:1执行完毕当前线程:2执行完毕// 到达信号量最大容量,其他线程就进行等待(阻塞)当前线程:3开始执行...当前线程:4开始执行...当前线程:3执行完毕当前线程:4执行完毕ReadWriteLock
主要使用到:ReentrantReadWriteLock(实现类)
概念
读写锁共存
独占/共享锁
//开启两个读写线程,分别进行写和读操作 for (int i = 0; i < 5; i++) { final Integer temp = i; new Thread(()->{ mapDemo.put(String.valueOf(temp), temp + 10000); }, "线程->" + String.valueOf(temp)).start(); } for (int i = 5; i < 10; i++) { final Integer temp = i; new Thread(()->{ mapDemo.get(String.valueOf(temp)); }, "线程->" + String.valueOf(temp)).start(); }// 初始化读写锁 private volatile Map<String, Object> map = new ConcurrentHashMap<>(); private ReadWriteLock readWriteLock = new ReentrantReadWriteLock(); public void put(String key, Object value) { //写入加锁 readWriteLock.writeLock().lock(); try { System.out.println(Thread.currentThread().getName() + "开始写入....."); map.put(key, value); System.out.println(Thread.currentThread().getName() + "写入完毕....."); }finally { //写完释放锁 readWriteLock.writeLock().unlock(); } } public Object get(String key) { //读取加锁 readWriteLock.readLock().lock(); Object object = null; try { System.out.println(Thread.currentThread().getName() + "开始读取----------->"); object = map.get(key); System.out.println(Thread.currentThread().getName() + "读取完成----------->"); }finally { //读取解锁 readWriteLock.readLock().unlock(); } return object; }// 运行结果:发现写入的时候,总是只有一个线程可以在同一时间进行写入,而读取可以多个线程同时读取线程->1开始写入.....线程->1写入完毕.....线程->0开始写入.....线程->0写入完毕.....线程->3开始写入.....线程->3写入完毕.....线程->2开始写入.....线程->2写入完毕.....线程->4开始写入.....线程->4写入完毕.....线程->5开始读取----------->线程->5读取完成----------->线程->7开始读取----------->线程->8开始读取----------->线程->8读取完成----------->线程->6开始读取----------->线程->9开始读取----------->线程->9读取完成----------->线程->7读取完成----------->线程->6读取完成-----------> ArrayBlockingQueue arrayBlockingQueue = new ArrayBlockingQueue<>(4); arrayBlockingQueue.add("A"); arrayBlockingQueue.add("B"); arrayBlockingQueue.add("C"); arrayBlockingQueue.add("D");// arrayBlockingQueue.add("E"); System.out.println(arrayBlockingQueue);//结果:[A, B, C, D]/**注意:1. 当元素超过队列的容量时,就会抛出异常java.lang.IllegalStateException: Queue full2. 当添加null时,抛出空指针异常 java.lang.NullPointerException3.使用offer代替add使用*///1.错误案例:容量为4,但是添加了五个元素 ArrayBlockingQueue arrayBlockingQueue = new ArrayBlockingQueue<>(4); arrayBlockingQueue.add("A"); arrayBlockingQueue.add("B"); arrayBlockingQueue.add("C"); arrayBlockingQueue.add("D"); arrayBlockingQueue.add("E"); System.out.println(arrayBlockingQueue);//结果:Exception in thread "main" java.lang.IllegalStateException: Queue full//2.错误案例:添加null数据,抛出空指针异常 ArrayBlockingQueue arrayBlockingQueue = new ArrayBlockingQueue<>(4); arrayBlockingQueue.add("A"); arrayBlockingQueue.add("B"); arrayBlockingQueue.add("C"); arrayBlockingQueue.add(null); System.out.println(arrayBlockingQueue);//结果:Exception in thread "main" java.lang.NullPointerException //3.使用offer代替add添加元素ArrayBlockingQueue arrayBlockingQueue = new ArrayBlockingQueue<>(4); arrayBlockingQueue.offer("A"); arrayBlockingQueue.offer("B"); arrayBlockingQueue.offer("C"); arrayBlockingQueue.offer("D"); arrayBlockingQueue.offer("E"); System.out.println(arrayBlockingQueue);//add与offer区别:add在超出容量时会抛出异常,而offer则不会抛出异常,而是拒绝添加到队列中!//结果:[A, B, C, D]ArrayBlockingQueue 延迟等待
//延迟添加等待----------------> offer()ArrayBlockingQueue arrayBlockingQueue = new ArrayBlockingQueue<>(4);System.out.println(arrayBlockingQueue.offer("A"));System.out.println(arrayBlockingQueue.offer("B"));System.out.println(arrayBlockingQueue.offer("C"));System.out.println(arrayBlockingQueue.offer("D"));//延迟12秒添加,如果队列已满就返回false(表示添加失败)System.out.println(arrayBlockingQueue.offer("E", 12, TimeUnit.SECONDS));System.out.println(arrayBlockingQueue);//结果:truetruetruetruefalse[A, B, C, D] // 延迟取出等待-------------> poll() ArrayBlockingQueue arrayBlockingQueue = new ArrayBlockingQueue<>(4); arrayBlockingQueue.offer("A"); arrayBlockingQueue.offer("B"); arrayBlockingQueue.offer("C"); arrayBlockingQueue.offer("D"); arrayBlockingQueue.offer("E", 2, TimeUnit.SECONDS); System.out.println(arrayBlockingQueue.poll()); System.out.println(arrayBlockingQueue.poll()); System.out.println(arrayBlockingQueue.poll()); System.out.println(arrayBlockingQueue.poll()); System.out.println(arrayBlockingQueue.poll(2, TimeUnit.SECONDS));// 结果:ABCD//延迟等待2秒钟再弹出,如果队列为空,就返回nullnull同步队列(SynchronousQueue)
存入之后必须等待取出之后才能进行再次存入栗子:
new Thread(()->{ try { synchronousQueue.put(1); System.out.println(Thread.currentThread().getName() + ":put " + 1); synchronousQueue.put(2); System.out.println(Thread.currentThread().getName() + ":put " + 2); synchronousQueue.put(3); System.out.println(Thread.currentThread().getName() + ":put " + 3); } catch (InterruptedException e) { e.printStackTrace(); } } , "put线程:").start();new Thread(()->{ try { System.out.println(Thread.currentThread().getName() + ":take -> " + synchronousQueue.take()); System.out.println(Thread.currentThread().getName() + ":take -> " + synchronousQueue.take()); System.out.println(Thread.currentThread().getName() + ":take -> " + synchronousQueue.take()); } catch (InterruptedException e) { e.printStackTrace(); } } , "take线程:").start();//结果:put线程::put 1take线程::take -> 1put线程::put 2take线程::take -> 2put线程::put 3take线程::take -> 3线程复用(节约了系统资源)
控制最大并发数(当达到线程池容量,就需要等待其他线程完成,才能继续进入)
管理线程
Executors线程池
ExecutorService executorService = Executors.newSingleThreadExecutor(); // 单个线程的池子ExecutorService executorService = Executors.newFixedThreadPool(10); //开启十个固定线程的池子ExecutorService executorService = Executors.newCachedThreadPool(); //可伸缩线程池, 如果线程池中线程已全被使用就创建新的线程池newSingleThreadExecutor
ExecutorService executorService = Executors.newSingleThreadExecutor(); // 单个线程的池子try { for (int i1 = 0; i1 < 10; i1++) { //执行线程 executorService.execute(new Runnable() { @Override public void run() { System.out.println(Thread.currentThread().getName() + " 执行了线程.."); } }); }} catch (Exception e) { e.printStackTrace();} finally { //关闭线程池 executorService.shutdown();}//结果:只有一个线程在重复利用执行pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-1 执行了线程..newFixedThreadPool
ExecutorService executorService = Executors.newFixedThreadPool(5); //开启十个固定线程的池子try { for (int i = 0; i < 10; i++) { //执行线程 executorService.execute(new Runnable() { @Override public void run() { System.out.println(Thread.currentThread().getName() + " 执行了线程.."); } }); }} catch (Exception e) { e.printStackTrace();} finally { //关闭线程池 executorService.shutdown();}//结果:五个不同的线程重复使用pool-1-thread-4 执行了线程..pool-1-thread-1 执行了线程..pool-1-thread-4 执行了线程..pool-1-thread-3 执行了线程..pool-1-thread-2 执行了线程..pool-1-thread-5 执行了线程..pool-1-thread-2 执行了线程..pool-1-thread-3 执行了线程..pool-1-thread-4 执行了线程..pool-1-thread-1 执行了线程..newCachedThreadPool
ExecutorService executorService = Executors.newCachedThreadPool(); //可伸缩线程池, 如果线程池中线程已全被使用就创建新的线程池try { for (int i = 0; i < 10; i++) { //开启线程 executorService.execute(new Runnable() { @Override public void run() { System.out.println(Thread.currentThread().getName() + " 执行了线程.."); } }); }} catch (Exception e) { e.printStackTrace();} finally { //关闭线程池 executorService.shutdown();}//结果:开启新线程,当已开启的线程执行完毕,放入池子中又可以进行使用,如果开启的线程都还在执行中,就创建新的线程pool-1-thread-1 执行了线程..pool-1-thread-6 执行了线程..pool-1-thread-5 执行了线程..pool-1-thread-3 执行了线程..pool-1-thread-4 执行了线程..pool-1-thread-2 执行了线程..pool-1-thread-8 执行了线程..pool-1-thread-7 执行了线程..pool-1-thread-9 执行了线程..pool-1-thread-10 执行了线程..线程池参数
public ThreadPoolExecutor(int corePoolSize, //核心线程数 int maximumPoolSize, //最大线程数 long keepAliveTime, //线程存活时间 TimeUnit unit, //线程时间单元 BlockingQueue<Runnable> workQueue) { //阻塞队列 this(corePoolSize, maximumPoolSize, keepAliveTime, unit, workQueue, Executors.defaultThreadFactory() //默认线程工厂 , defaultHandler); //线程拒绝策略(当达到了最大线程数时,采用线程拒绝策略)}Spring自带的任务执行器线程池
@Bean("scheduledTaskExecutor")public ThreadPoolTaskExecutor scheduledTaskExecutor() { ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor(); //创建任务执行器线程池 executor.setCorePoolSize(3);//设置核心线程数 executor.setMaxPoolSize(5); //设置最大线程数 executor.setQueueCapacity(1024*100); //设置一个队列容量 executor.setThreadNamePrefix("parking-index-task"); //线程名称 executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy()); // 拒绝策略 executor.initialize(); //初始化线程池 return executor;}线程四大拒绝策略应用场景
AbortPolicy: 当队列中线程已满,就抛出异常DiscardPolicy:当队列满了,就丢弃任务,不会抛出异常CallerRunsPolicy: 队列已满时,就使用调用者的线程去执行,当处理器关闭就丢弃此线程需求DiscardOldestPolicy:当队列满了,去尝试和较早的线程竞争,当最早的线程即将执行完成就把当前任务使用即将完成的线程执行源码解释:AbortPolicy:public AbortPolicy() { } /** * Always throws RejectedExecutionException. * * @param r the runnable task requested to be executed * @param e the executor attempting to execute this task * @throws RejectedExecutionException always 解释:总是把RejectedExecutionException。Params: r—请求执行的可运行任务e—尝试执行该任务的执行器抛出:RejectedExecutionException—always */ public void rejectedExecution(Runnable r, ThreadPoolExecutor e) { throw new RejectedExecutionException("Task " + r.toString() + " rejected from " + e.toString()); }DiscardPolicy: public DiscardPolicy() { } /** * Does nothing, which has the effect of discarding task r. * * @param r the runnable task requested to be executed * @param e the executor attempting to execute this task 解释:什么都不做,这有丢弃任务r的效果。参数:r -请求被执行的可运行任务e -试图执行该任务的执行程序 */ public void rejectedExecution(Runnable r, ThreadPoolExecutor e) { }CallerRunsPolicy: public CallerRunsPolicy() { } /** * Executes task r in the caller's thread, unless the executor * has been shut down, in which case the task is discarded. * * @param r the runnable task requested to be executed * @param e the executor attempting to execute this task 解释:在调用者的线程中执行任务r,除非执行器已经关闭,在这种情况下,任务将被丢弃。参数:r—请求执行的可运行任务e—尝试执行该任务的执行程序 */ public void rejectedExecution(Runnable r, ThreadPoolExecutor e) { if (!e.isShutdown()) { r.run(); }DiscardOldestPolicy: public DiscardOldestPolicy() { } /** * Obtains and ignores the next task that the executor * would otherwise execute, if one is immediately available, * and then retries execution of task r, unless the executor * is shut down, in which case task r is instead discarded. * * @param r the runnable task requested to be executed * @param e the executor attempting to execute this task 解释:获取并忽略执行器将执行的下一个任务(如果有一个任务立即可用),然后重试执行任务r,除非执行器被关闭,在这种情况下,任务r将被丢弃。参数:r—请求执行的可运行任务e—尝试执行该任务的执行程序 */ public void rejectedExecution(Runnable r, ThreadPoolExecutor e) { if (!e.isShutdown()) { e.getQueue().poll(); e.execute(r); }Cpu密集型
int availableProcessors = Runtime.getRuntime().availableProcessors(); //获取Cpu核数,适合设置核心线程池的大小IO密集型
int availableProcessors = Runtime.getRuntime().availableProcessors(); //int maximumPoolSize = availableProcessors * 2; // Io密集型一般设置为Cpu核数的两倍,防止public class DoMain extends RecursiveTask<Long> { private Long start; private Long end; private final Long threshold = 10_0000_0000L; public DoMain(Long start, Long end) { this.start = start; this.end = end; } /** 递归分解大数据,每次进行两段两段操作 */ @Override protected Long compute() { Long res = 0L; if ((end - start) > threshold) { Long middle = (end + start) / 2; //分两次进行计算 ForkJoinTask<Long> fork1 = new DoMain(start, middle).fork(); Long res1 = fork1.join(); ForkJoinTask<Long> fork2 = new DoMain(middle, end).fork(); Long res2 = fork2.join(); res = res1 + res2; } else { for (Long i = start; i < end; i++) { res += i; } } return res; }} //这样创建线程不规范,这里只是简易操作! new Thread(() -> { long l = System.currentTimeMillis(); DoMain doMain = new DoMain(0L, 500_0000_0000L); ForkJoinTask<Long> submit = new ForkJoinPool().submit(doMain); try { System.out.println("forkJoin输出结果:" + submit.get()); } catch (InterruptedException e) { e.printStackTrace(); } catch (ExecutionException e) { e.printStackTrace(); } System.out.println("forkJoin所用时间: " + (System.currentTimeMillis() - l)); }).start(); //这样创建线程不规范,这里只是简易操作! new Thread(() -> { long start = System.currentTimeMillis(); Long res = 0L; for (Long i = 0L; i < 500_0000_0000L; i++) { res += i; } System.out.println("普通循环输出结果:" + res); System.out.println("普通所用时间" + (System.currentTimeMillis() - start)); }).start();计算对比
stream环输出结果: 124999999750000000
stream所用时间2304
普通循环输出结果:124999999750000000
普通所用时间14116
forkJoin输出结果:124999999750000000
forkJoin所用时间: 14468
new Thread(() -> { long start = System.currentTimeMillis(); long longStream = LongStream.range(0L, 5_0000_0000L).parallel().reduce(0L, Long::sum); System.out.println("stream环输出结果: " + longStream); System.out.println("stream所用时间" + (System.currentTimeMillis() - start)); }).start();/** * 1.使用volatile禁止指令重排 * 2. 使用AtomicInteger原子类保证是原子操作 */public static volatile AtomicInteger num = new AtomicInteger();public static void main(String[] args) { for (int i = 0; i < 20; i++) { new Thread(() -> { for (int j = 0; j < 1000; j++) { //进行加一操作 num.getAndIncrement(); } }).start(); } //当线程数大于2时,暂停main线程,让给其他线程执行 while (Thread.activeCount() > 2) { Thread.yield(); } System.out.println(num);原子类操作源码
public final int getAndIncrement() { return unsafe.getAndAddInt(this, valueOffset, 1);} //原子类底层代码,使用了CAS(比较替换算法,也就是自旋锁) // compareAndSwapInt底层是调用c++操作内存,对应的是native关键字 public final int getAndAddInt(Object var1, long var2, int var4) { int var5; do { var5 = this.getIntVolatile(var1, var2); } while(!this.compareAndSwapInt(var1, var2, var5, var5 + var4)); return var5; }栗子:
@SneakyThrowspublic static void main(String[] args) { CasLock casLock = new CasLock(); new Thread(()->{ try { casLock.lock(); }catch (Exception e) { e.printStackTrace(); }finally { casLock.unLock(); } }, "Thread1").start(); TimeUnit.SECONDS.sleep(2); new Thread(()->{ try { casLock.lock(); }catch (Exception e) { e.printStackTrace(); }finally { casLock.unLock(); } }, "Thread2").start();}public static class CasLock { AtomicReference<Thread> lock = new AtomicReference<>(); public void lock() { Thread thread = Thread.currentThread(); if (lock.get() == null) { //拿到泛型中Thread的值进行比较 System.out.println(thread.getName() + "----> 开始自旋..."); } while (lock.compareAndSet(null, thread)) { } } public void unLock() { Thread thread = Thread.currentThread(); System.out.println(thread.getName() + "----> 解锁成功!"); //解锁 lock.compareAndSet(thread, null); }}