一、什么是死锁一个生动的比喻想象这个场景你在餐厅吃饭左手拿刀右手拿叉。 你朋友也来吃饭但他先拿走了叉然后伸手来拿你的刀。 你不给刀因为你需要叉才能吃饭。 他也不给叉因为他需要刀才能吃饭。结果你们俩都吃不上饭僵持到地老天荒。这就是死锁在多线程编程中当两个或多个线程互相等待对方释放资源时程序就会永远卡住。二、死锁的四大必要条件死锁不会随便发生必须同时满足四个条件条件1互斥Mutual Exclusion// 资源一次只能被一个线程使用 public class MutexExample { private final Object lock new Object(); public void exclusiveMethod() { synchronized (lock) { // 互斥访问 // 同一时间只能有一个线程进入 } } }条件2请求与保持Hold and Waitpublic class HoldAndWaitExample { private final Object lockA new Object(); private final Object lockB new Object(); public void problematicMethod() { synchronized (lockA) { // 持有lockA // ... 一些操作 synchronized (lockB) { // 请求lockB // 危险持有lockA的同时请求lockB } } } }条件3不可剥夺No Preemption// 线程已获得的资源不能被强制剥夺 // 在Java中除非线程主动释放否则锁会一直被持有 public class NoPreemptionExample { public void dangerousMethod(Object lock) { synchronized (lock) { // 除非这个代码块执行完毕或者线程被中断 // 否则其他线程无法强制获取这个锁 // 即使是高优先级的线程也不行 } } }条件4循环等待Circular Wait// 经典死锁示例 public class CircularWaitExample { private static final Object lock1 new Object(); private static final Object lock2 new Object(); public static void main(String[] args) { // 线程1先获取lock1再请求lock2 Thread t1 new Thread(() - { synchronized (lock1) { System.out.println(线程1获取lock1); try { Thread.sleep(100); } catch (InterruptedException e) {} synchronized (lock2) { // 等待线程2释放lock2 System.out.println(线程1获取lock2); } } }); // 线程2先获取lock2再请求lock1 Thread t2 new Thread(() - { synchronized (lock2) { System.out.println(线程2获取lock2); try { Thread.sleep(100); } catch (InterruptedException e) {} synchronized (lock1) { // 等待线程1释放lock1 System.out.println(线程2获取lock1); } } }); t1.start(); t2.start(); // 结果两个线程永远等待形成循环等待链 // 线程1 - 等待lock2 - 被线程2持有 // 线程2 - 等待lock1 - 被线程1持有 } }三、真实案例银行转账死锁让我们看一个经典的银行转账死锁案例// ❌ 错误实现致命的死锁陷阱 public class BankAccount { private String accountId; private BigDecimal balance; // 转账方法 - 存在死锁风险 public static void transfer(BankAccount from, BankAccount to, BigDecimal amount) { synchronized (from) { // 获取转出账户锁 synchronized (to) { // 获取转入账户锁 if (from.getBalance().compareTo(amount) 0) { from.withdraw(amount); to.deposit(amount); System.out.println(转账成功: amount); } } } } // 问题来了如果同时发生两笔转账 // 转账1: A - B // 转账2: B - A // 线程1: 锁定A等待B // 线程2: 锁定B等待A // 死锁 }四、5分钟快速诊断你的程序死锁了吗第一步获取线程转储# Linux/Mac jstackthread_dump.txt # 或者使用jcmd jcmdThread.print # 如果在容器中 kubectl exec --jstack 1dump.txt第二步查找死锁关键字在线程转储中搜索这些关键词Found one Java-level deadlock: # 发现死锁 Thread-1: waiting to lock monitor 0x00007f8a4c0c3b58 (object 0x00000000f6f8f9b8) which is held by Thread-2 Thread-2: waiting to lock monitor 0x00007f8a4c0c3c58 (object 0x00000000f6f8f9a8) which is held by Thread-1第三步使用JConsole/JVisualVM可视化分析!https://example.com/deadlock-detection.png图形化工具可以直观显示死锁关系五、死锁解决三板斧第一斧避免锁顺序不一致最常用// ✅ 正确实现通过统一锁顺序避免死锁 public class BankAccountSafe { private String accountId; private BigDecimal balance; public String getAccountId() { return accountId; } // 安全的转账实现 public static void transferSafe(BankAccountSafe from, BankAccountSafe to, BigDecimal amount) { // 关键确定全局的锁顺序 BankAccountSafe firstLock from; BankAccountSafe secondLock to; // 通过账户ID确定锁定顺序 if (from.getAccountId().compareTo(to.getAccountId()) 0) { firstLock to; secondLock from; } synchronized (firstLock) { synchronized (secondLock) { if (from.getBalance().compareTo(amount) 0) { from.withdraw(amount); to.deposit(amount); System.out.println(安全转账: amount); } } } } }第二斧使用tryLock超时机制import java.util.concurrent.TimeUnit; import java.util.concurrent.locks.Lock; import java.util.concurrent.locks.ReentrantLock; public class TryLockSolution { private final Lock lock1 new ReentrantLock(); private final Lock lock2 new ReentrantLock(); public boolean tryTransfer(long timeout, TimeUnit unit) throws InterruptedException { long stopTime System.nanoTime() unit.toNanos(timeout); while (true) { // 尝试获取第一把锁 if (lock1.tryLock()) { try { // 尝试获取第二把锁 if (lock2.tryLock()) { try { // 执行业务逻辑 doBusiness(); return true; } finally { lock2.unlock(); } } } finally { lock1.unlock(); // 释放第一把锁 } } // 检查是否超时 if (System.nanoTime() stopTime) { return false; // 超时返回失败 } // 短暂休眠避免CPU忙等 Thread.sleep(new Random().nextInt(10)); } } }第三斧使用资源层级锁public class ResourceHierarchy { // 定义全局资源层级 private static final int RESOURCE_A_LEVEL 1; private static final int RESOURCE_B_LEVEL 2; private static final int RESOURCE_C_LEVEL 3; public void accessResources(Object resourceA, Object resourceB, Object resourceC) { // 按照资源层级顺序加锁 Object level1 getResourceByLevel(RESOURCE_A_LEVEL, resourceA, resourceB, resourceC); Object level2 getResourceByLevel(RESOURCE_B_LEVEL, resourceA, resourceB, resourceC); Object level3 getResourceByLevel(RESOURCE_C_LEVEL, resourceA, resourceB, resourceC); synchronized (level1) { synchronized (level2) { synchronized (level3) { // 安全地访问所有资源 } } } } }六、高级技巧死锁检测与自动恢复1. 使用守护线程监控Component Slf4j public class DeadlockDetector { private final ScheduledExecutorService scheduler Executors.newSingleThreadScheduledExecutor(); PostConstruct public void startDetection() { // 每30秒检测一次死锁 scheduler.scheduleAtFixedRate(this::detectAndResolve, 0, 30, TimeUnit.SECONDS); } private void detectAndResolve() { ThreadMXBean threadBean ManagementFactory.getThreadMXBean(); // 查找死锁线程 long[] deadlockedThreadIds threadBean.findDeadlockedThreads(); if (deadlockedThreadIds ! null deadlockedThreadIds.length 0) { log.error(发现死锁涉及线程数: {}, deadlockedThreadIds.length); // 记录死锁详情 for (long threadId : deadlockedThreadIds) { ThreadInfo threadInfo threadBean.getThreadInfo(threadId); log.error(死锁线程: {}, 状态: {}, 等待锁: {}, 持有锁: {}, threadInfo.getThreadName(), threadInfo.getThreadState(), threadInfo.getLockName(), threadInfo.getLockOwnerName()); } // 尝试自动恢复谨慎使用 autoRecover(deadlockedThreadIds); } } private void autoRecover(long[] deadlockedThreadIds) { // 策略1中断其中一个死锁线程 ThreadInfo[] threadInfos ManagementFactory.getThreadMXBean() .getThreadInfo(deadlockedThreadIds, 0); if (threadInfos.length 0) { // 找出所有线程选择优先级最低的进行中断 Thread targetThread findThreadById(threadInfos[0].getThreadId()); if (targetThread ! null) { log.warn(尝试中断线程以解除死锁: {}, targetThread.getName()); targetThread.interrupt(); // 发送告警通知 sendAlert(系统检测到死锁已自动中断线程: targetThread.getName()); } } } }2. 使用断路器模式Slf4j public class CircuitBreakerWithDeadlockProtection { private final Lock lock new ReentrantLock(); private final AtomicInteger failureCount new AtomicInteger(0); private volatile long lastFailureTime 0; private volatile State state State.CLOSED; enum State { CLOSED, OPEN, HALF_OPEN } public T T execute(SupplierT supplier, T fallbackValue) { if (state State.OPEN) { if (System.currentTimeMillis() - lastFailureTime 5000) { state State.HALF_OPEN; } else { return fallbackValue; // 快速失败 } } if (!lock.tryLock()) { // 如果无法快速获取锁可能发生死锁风险 failureCount.incrementAndGet(); if (failureCount.get() 10) { state State.OPEN; // 打开断路器 lastFailureTime System.currentTimeMillis(); } return fallbackValue; } try { T result supplier.get(); // 成功重置计数器 failureCount.set(0); state State.CLOSED; return result; } finally { lock.unlock(); } } }七、生产环境死锁事故复盘事故背景某电商平台在双11大促期间订单系统突然卡死所有下单请求超时。问题代码Service public class OrderService { Autowired private InventoryService inventoryService; Autowired private CouponService couponService; Transactional public Order createOrder(OrderRequest request) { // 锁定库存 synchronized (inventoryService.getLock(request.getProductId())) { // 锁定优惠券 synchronized (couponService.getLock(request.getUserId())) { // 业务逻辑 } } } public void cancelOrder(String orderId) { // 先锁定优惠券 synchronized (couponService.getLock(userId)) { // 再锁定库存 synchronized (inventoryService.getLock(productId)) { // 业务逻辑 } } } }死锁发生场景用户A下单锁定库存A → 请求锁定优惠券A同时用户A取消订单锁定优惠券A → 请求锁定库存A结果创建订单线程和取消订单线程互相等待死锁解决方案// ✅ 修复方案统一锁顺序 Service public class OrderServiceFixed { // 定义全局锁顺序先锁用户再锁商品 public void processOrder(OrderRequest request, boolean isCreate) { Object userLock getUserLock(request.getUserId()); Object productLock getProductLock(request.getProductId()); // 统一先锁用户再锁商品 synchronized (userLock) { synchronized (productLock) { if (isCreate) { createOrderInternal(request); } else { cancelOrderInternal(request); } } } } // 使用数据库行锁替代synchronized Transactional public Order createOrderWithRowLock(OrderRequest request) { // 使用SELECT ... FOR UPDATE 锁定用户记录 User user userRepository.findByIdForUpdate(request.getUserId()); // 使用SELECT ... FOR UPDATE 锁定商品记录 Product product productRepository.findByIdForUpdate(request.getProductId()); // 业务逻辑... } }八、预防死锁的铁律铁律1锁顺序一致性// ❌ 错误不同方法中锁顺序不一致 public void method1() { sync(A) { sync(B) {} } } public void method2() { sync(B) { sync(A) {} } } // 危险 // ✅ 正确全局统一定义锁顺序 private static final ComparatorObject LOCK_ORDER (o1, o2) - System.identityHashCode(o1) - System.identityHashCode(o2); public void safeMethod(Object a, Object b) { Object first a, second b; if (LOCK_ORDER.compare(a, b) 0) { first b; second a; } synchronized (first) { synchronized (second) { // ... } } }铁律2锁超时机制// 使用ReentrantLock的tryLock private final Lock lock new ReentrantLock(); public void doWithTimeout(long timeout, TimeUnit unit) { try { if (lock.tryLock(timeout, unit)) { try { // 业务逻辑 } finally { lock.unlock(); } } else { // 超时处理记录告警降级处理 log.warn(获取锁超时执行降级逻辑); fallback(); } } catch (InterruptedException e) { Thread.currentThread().interrupt(); } }铁律3减少锁粒度// ❌ 错误锁粒度过大 public synchronized void processOrder() { // 锁整个对象 // 几十行代码... } // ✅ 正确细化锁粒度 private final Object orderLock new Object(); private final Object userLock new Object(); public void processOrder() { // 只有需要同步的部分加锁 synchronized (orderLock) { // 仅同步订单相关操作 } // 其他非同步操作 updateLog(); }铁律4使用无锁数据结构// 使用并发集合代替手动同步 private final ConcurrentHashMapString, User userCache new ConcurrentHashMap(); private final AtomicInteger counter new AtomicInteger(0); private final LongAdder totalAmount new LongAdder(); // 使用CopyOnWriteArrayList避免读锁 private final CopyOnWriteArrayListListener listeners new CopyOnWriteArrayList();九、死锁检测工具推荐VisualVM- 图形化死锁检测JConsole- JDK自带监控工具arthas- 阿里开源的Java诊断工具jstack- 命令行线程转储分析YourKit- 商业级性能分析工具# 使用arthas检测死锁 $java-jararthas-boot.jar $dashboard # 查看实时监控 $thread-b # 检测死锁 $thread--stateBLOCKED # 查看阻塞线程十、总结死锁防御体系构建完善的死锁防御体系需要从四个层面入手1. 编码规范层制定团队锁使用规范代码审查重点检查锁顺序使用静态代码分析工具如SonarQube2. 架构设计层避免过度使用同步采用消息队列解耦使用乐观锁、无锁编程3. 监控告警层实时监控线程状态死锁自动检测告警定期生成线程分析报告4. 应急预案层制定死锁应急处理流程准备熔断降级方案建立自动恢复机制记住预防死锁比解决死锁更重要。良好的设计和规范能让你的系统远离鬼打墙的困境。最后的小测试看看下面这段代码有没有死锁风险如何修复public class TestDeadlock { private static final Object lock1 new Object(); private static final Object lock2 new Object(); public static void main(String[] args) { new Thread(() - { synchronized (lock1) { synchronized (lock2) { System.out.println(Thread1); } } }).start(); new Thread(() - { synchronized (lock2) { synchronized (lock1) { System.out.println(Thread2); } } }).start(); } }把你的答案写在评论区我们一起学习
线程死锁：程序员的“鬼打墙“，如何破解？
一、什么是死锁一个生动的比喻想象这个场景你在餐厅吃饭左手拿刀右手拿叉。 你朋友也来吃饭但他先拿走了叉然后伸手来拿你的刀。 你不给刀因为你需要叉才能吃饭。 他也不给叉因为他需要刀才能吃饭。结果你们俩都吃不上饭僵持到地老天荒。这就是死锁在多线程编程中当两个或多个线程互相等待对方释放资源时程序就会永远卡住。二、死锁的四大必要条件死锁不会随便发生必须同时满足四个条件条件1互斥Mutual Exclusion// 资源一次只能被一个线程使用 public class MutexExample { private final Object lock new Object(); public void exclusiveMethod() { synchronized (lock) { // 互斥访问 // 同一时间只能有一个线程进入 } } }条件2请求与保持Hold and Waitpublic class HoldAndWaitExample { private final Object lockA new Object(); private final Object lockB new Object(); public void problematicMethod() { synchronized (lockA) { // 持有lockA // ... 一些操作 synchronized (lockB) { // 请求lockB // 危险持有lockA的同时请求lockB } } } }条件3不可剥夺No Preemption// 线程已获得的资源不能被强制剥夺 // 在Java中除非线程主动释放否则锁会一直被持有 public class NoPreemptionExample { public void dangerousMethod(Object lock) { synchronized (lock) { // 除非这个代码块执行完毕或者线程被中断 // 否则其他线程无法强制获取这个锁 // 即使是高优先级的线程也不行 } } }条件4循环等待Circular Wait// 经典死锁示例 public class CircularWaitExample { private static final Object lock1 new Object(); private static final Object lock2 new Object(); public static void main(String[] args) { // 线程1先获取lock1再请求lock2 Thread t1 new Thread(() - { synchronized (lock1) { System.out.println(线程1获取lock1); try { Thread.sleep(100); } catch (InterruptedException e) {} synchronized (lock2) { // 等待线程2释放lock2 System.out.println(线程1获取lock2); } } }); // 线程2先获取lock2再请求lock1 Thread t2 new Thread(() - { synchronized (lock2) { System.out.println(线程2获取lock2); try { Thread.sleep(100); } catch (InterruptedException e) {} synchronized (lock1) { // 等待线程1释放lock1 System.out.println(线程2获取lock1); } } }); t1.start(); t2.start(); // 结果两个线程永远等待形成循环等待链 // 线程1 - 等待lock2 - 被线程2持有 // 线程2 - 等待lock1 - 被线程1持有 } }三、真实案例银行转账死锁让我们看一个经典的银行转账死锁案例// ❌ 错误实现致命的死锁陷阱 public class BankAccount { private String accountId; private BigDecimal balance; // 转账方法 - 存在死锁风险 public static void transfer(BankAccount from, BankAccount to, BigDecimal amount) { synchronized (from) { // 获取转出账户锁 synchronized (to) { // 获取转入账户锁 if (from.getBalance().compareTo(amount) 0) { from.withdraw(amount); to.deposit(amount); System.out.println(转账成功: amount); } } } } // 问题来了如果同时发生两笔转账 // 转账1: A - B // 转账2: B - A // 线程1: 锁定A等待B // 线程2: 锁定B等待A // 死锁 }四、5分钟快速诊断你的程序死锁了吗第一步获取线程转储# Linux/Mac jstackthread_dump.txt # 或者使用jcmd jcmdThread.print # 如果在容器中 kubectl exec --jstack 1dump.txt第二步查找死锁关键字在线程转储中搜索这些关键词Found one Java-level deadlock: # 发现死锁 Thread-1: waiting to lock monitor 0x00007f8a4c0c3b58 (object 0x00000000f6f8f9b8) which is held by Thread-2 Thread-2: waiting to lock monitor 0x00007f8a4c0c3c58 (object 0x00000000f6f8f9a8) which is held by Thread-1第三步使用JConsole/JVisualVM可视化分析!https://example.com/deadlock-detection.png图形化工具可以直观显示死锁关系五、死锁解决三板斧第一斧避免锁顺序不一致最常用// ✅ 正确实现通过统一锁顺序避免死锁 public class BankAccountSafe { private String accountId; private BigDecimal balance; public String getAccountId() { return accountId; } // 安全的转账实现 public static void transferSafe(BankAccountSafe from, BankAccountSafe to, BigDecimal amount) { // 关键确定全局的锁顺序 BankAccountSafe firstLock from; BankAccountSafe secondLock to; // 通过账户ID确定锁定顺序 if (from.getAccountId().compareTo(to.getAccountId()) 0) { firstLock to; secondLock from; } synchronized (firstLock) { synchronized (secondLock) { if (from.getBalance().compareTo(amount) 0) { from.withdraw(amount); to.deposit(amount); System.out.println(安全转账: amount); } } } } }第二斧使用tryLock超时机制import java.util.concurrent.TimeUnit; import java.util.concurrent.locks.Lock; import java.util.concurrent.locks.ReentrantLock; public class TryLockSolution { private final Lock lock1 new ReentrantLock(); private final Lock lock2 new ReentrantLock(); public boolean tryTransfer(long timeout, TimeUnit unit) throws InterruptedException { long stopTime System.nanoTime() unit.toNanos(timeout); while (true) { // 尝试获取第一把锁 if (lock1.tryLock()) { try { // 尝试获取第二把锁 if (lock2.tryLock()) { try { // 执行业务逻辑 doBusiness(); return true; } finally { lock2.unlock(); } } } finally { lock1.unlock(); // 释放第一把锁 } } // 检查是否超时 if (System.nanoTime() stopTime) { return false; // 超时返回失败 } // 短暂休眠避免CPU忙等 Thread.sleep(new Random().nextInt(10)); } } }第三斧使用资源层级锁public class ResourceHierarchy { // 定义全局资源层级 private static final int RESOURCE_A_LEVEL 1; private static final int RESOURCE_B_LEVEL 2; private static final int RESOURCE_C_LEVEL 3; public void accessResources(Object resourceA, Object resourceB, Object resourceC) { // 按照资源层级顺序加锁 Object level1 getResourceByLevel(RESOURCE_A_LEVEL, resourceA, resourceB, resourceC); Object level2 getResourceByLevel(RESOURCE_B_LEVEL, resourceA, resourceB, resourceC); Object level3 getResourceByLevel(RESOURCE_C_LEVEL, resourceA, resourceB, resourceC); synchronized (level1) { synchronized (level2) { synchronized (level3) { // 安全地访问所有资源 } } } } }六、高级技巧死锁检测与自动恢复1. 使用守护线程监控Component Slf4j public class DeadlockDetector { private final ScheduledExecutorService scheduler Executors.newSingleThreadScheduledExecutor(); PostConstruct public void startDetection() { // 每30秒检测一次死锁 scheduler.scheduleAtFixedRate(this::detectAndResolve, 0, 30, TimeUnit.SECONDS); } private void detectAndResolve() { ThreadMXBean threadBean ManagementFactory.getThreadMXBean(); // 查找死锁线程 long[] deadlockedThreadIds threadBean.findDeadlockedThreads(); if (deadlockedThreadIds ! null deadlockedThreadIds.length 0) { log.error(发现死锁涉及线程数: {}, deadlockedThreadIds.length); // 记录死锁详情 for (long threadId : deadlockedThreadIds) { ThreadInfo threadInfo threadBean.getThreadInfo(threadId); log.error(死锁线程: {}, 状态: {}, 等待锁: {}, 持有锁: {}, threadInfo.getThreadName(), threadInfo.getThreadState(), threadInfo.getLockName(), threadInfo.getLockOwnerName()); } // 尝试自动恢复谨慎使用 autoRecover(deadlockedThreadIds); } } private void autoRecover(long[] deadlockedThreadIds) { // 策略1中断其中一个死锁线程 ThreadInfo[] threadInfos ManagementFactory.getThreadMXBean() .getThreadInfo(deadlockedThreadIds, 0); if (threadInfos.length 0) { // 找出所有线程选择优先级最低的进行中断 Thread targetThread findThreadById(threadInfos[0].getThreadId()); if (targetThread ! null) { log.warn(尝试中断线程以解除死锁: {}, targetThread.getName()); targetThread.interrupt(); // 发送告警通知 sendAlert(系统检测到死锁已自动中断线程: targetThread.getName()); } } } }2. 使用断路器模式Slf4j public class CircuitBreakerWithDeadlockProtection { private final Lock lock new ReentrantLock(); private final AtomicInteger failureCount new AtomicInteger(0); private volatile long lastFailureTime 0; private volatile State state State.CLOSED; enum State { CLOSED, OPEN, HALF_OPEN } public T T execute(SupplierT supplier, T fallbackValue) { if (state State.OPEN) { if (System.currentTimeMillis() - lastFailureTime 5000) { state State.HALF_OPEN; } else { return fallbackValue; // 快速失败 } } if (!lock.tryLock()) { // 如果无法快速获取锁可能发生死锁风险 failureCount.incrementAndGet(); if (failureCount.get() 10) { state State.OPEN; // 打开断路器 lastFailureTime System.currentTimeMillis(); } return fallbackValue; } try { T result supplier.get(); // 成功重置计数器 failureCount.set(0); state State.CLOSED; return result; } finally { lock.unlock(); } } }七、生产环境死锁事故复盘事故背景某电商平台在双11大促期间订单系统突然卡死所有下单请求超时。问题代码Service public class OrderService { Autowired private InventoryService inventoryService; Autowired private CouponService couponService; Transactional public Order createOrder(OrderRequest request) { // 锁定库存 synchronized (inventoryService.getLock(request.getProductId())) { // 锁定优惠券 synchronized (couponService.getLock(request.getUserId())) { // 业务逻辑 } } } public void cancelOrder(String orderId) { // 先锁定优惠券 synchronized (couponService.getLock(userId)) { // 再锁定库存 synchronized (inventoryService.getLock(productId)) { // 业务逻辑 } } } }死锁发生场景用户A下单锁定库存A → 请求锁定优惠券A同时用户A取消订单锁定优惠券A → 请求锁定库存A结果创建订单线程和取消订单线程互相等待死锁解决方案// ✅ 修复方案统一锁顺序 Service public class OrderServiceFixed { // 定义全局锁顺序先锁用户再锁商品 public void processOrder(OrderRequest request, boolean isCreate) { Object userLock getUserLock(request.getUserId()); Object productLock getProductLock(request.getProductId()); // 统一先锁用户再锁商品 synchronized (userLock) { synchronized (productLock) { if (isCreate) { createOrderInternal(request); } else { cancelOrderInternal(request); } } } } // 使用数据库行锁替代synchronized Transactional public Order createOrderWithRowLock(OrderRequest request) { // 使用SELECT ... FOR UPDATE 锁定用户记录 User user userRepository.findByIdForUpdate(request.getUserId()); // 使用SELECT ... FOR UPDATE 锁定商品记录 Product product productRepository.findByIdForUpdate(request.getProductId()); // 业务逻辑... } }八、预防死锁的铁律铁律1锁顺序一致性// ❌ 错误不同方法中锁顺序不一致 public void method1() { sync(A) { sync(B) {} } } public void method2() { sync(B) { sync(A) {} } } // 危险 // ✅ 正确全局统一定义锁顺序 private static final ComparatorObject LOCK_ORDER (o1, o2) - System.identityHashCode(o1) - System.identityHashCode(o2); public void safeMethod(Object a, Object b) { Object first a, second b; if (LOCK_ORDER.compare(a, b) 0) { first b; second a; } synchronized (first) { synchronized (second) { // ... } } }铁律2锁超时机制// 使用ReentrantLock的tryLock private final Lock lock new ReentrantLock(); public void doWithTimeout(long timeout, TimeUnit unit) { try { if (lock.tryLock(timeout, unit)) { try { // 业务逻辑 } finally { lock.unlock(); } } else { // 超时处理记录告警降级处理 log.warn(获取锁超时执行降级逻辑); fallback(); } } catch (InterruptedException e) { Thread.currentThread().interrupt(); } }铁律3减少锁粒度// ❌ 错误锁粒度过大 public synchronized void processOrder() { // 锁整个对象 // 几十行代码... } // ✅ 正确细化锁粒度 private final Object orderLock new Object(); private final Object userLock new Object(); public void processOrder() { // 只有需要同步的部分加锁 synchronized (orderLock) { // 仅同步订单相关操作 } // 其他非同步操作 updateLog(); }铁律4使用无锁数据结构// 使用并发集合代替手动同步 private final ConcurrentHashMapString, User userCache new ConcurrentHashMap(); private final AtomicInteger counter new AtomicInteger(0); private final LongAdder totalAmount new LongAdder(); // 使用CopyOnWriteArrayList避免读锁 private final CopyOnWriteArrayListListener listeners new CopyOnWriteArrayList();九、死锁检测工具推荐VisualVM- 图形化死锁检测JConsole- JDK自带监控工具arthas- 阿里开源的Java诊断工具jstack- 命令行线程转储分析YourKit- 商业级性能分析工具# 使用arthas检测死锁 $java-jararthas-boot.jar $dashboard # 查看实时监控 $thread-b # 检测死锁 $thread--stateBLOCKED # 查看阻塞线程十、总结死锁防御体系构建完善的死锁防御体系需要从四个层面入手1. 编码规范层制定团队锁使用规范代码审查重点检查锁顺序使用静态代码分析工具如SonarQube2. 架构设计层避免过度使用同步采用消息队列解耦使用乐观锁、无锁编程3. 监控告警层实时监控线程状态死锁自动检测告警定期生成线程分析报告4. 应急预案层制定死锁应急处理流程准备熔断降级方案建立自动恢复机制记住预防死锁比解决死锁更重要。良好的设计和规范能让你的系统远离鬼打墙的困境。最后的小测试看看下面这段代码有没有死锁风险如何修复public class TestDeadlock { private static final Object lock1 new Object(); private static final Object lock2 new Object(); public static void main(String[] args) { new Thread(() - { synchronized (lock1) { synchronized (lock2) { System.out.println(Thread1); } } }).start(); new Thread(() - { synchronized (lock2) { synchronized (lock1) { System.out.println(Thread2); } } }).start(); } }把你的答案写在评论区我们一起学习
