Require：基于雪花算法完成一个局部随机，全局离散没有热点切唯一的数值Id生成器。
1，雪花算法
【雪花算法】雪花算法（Snowflake Algorithm）是Twitter开源的用于生成唯一ID的算法，它可以在分布式系统中生成唯一的64位长整数ID。这种ID生成方式既保证了趋势递增，又保证了在不同数据中心、不同机器上生成的ID的唯一性。
符号位：占用1位。
时间戳：通常占用41位，表示从某个固定时间点（如1970年1月1日）起的毫秒数。这使得生成的ID具有时间排序性。
机器ID：占用10位，表示生成ID的机器节点，以确保不同机器生成的ID不冲突。
序列号：占用12位，允许同一毫秒内生成多个ID，通常用于支持高并发。
public class SnowflakeIdGenerator { private static final long EPOCH = 1640995200000L; // 自定义的开始时间戳（2022年1月1日） private static final long MACHINE_ID_BITS = 10L; // 机器ID的位数 private static final long SEQUENCE_BITS = 12L; // 序列号的位数 private static final long MAX_MACHINE_ID = ~(-1L << MACHINE_ID_BITS); // 机器ID最大值 private static final long SEQUENCE_MASK = ~(-1L << SEQUENCE_BITS); // 序列号最大值 private long machineId; // 当前机器ID private long sequence = 0L; // 当前序列号 private long lastTimestamp = -1L; // 上次生成ID的时间戳 public SnowflakeIdGenerator(long machineId) { if (machineId > MAX_MACHINE_ID || machineId < 0) { throw new IllegalArgumentException("Machine ID can't be greater than " + MAX_MACHINE_ID + " or less than 0"); } this.machineId = machineId; } public synchronized long generateId() { long timestamp = System.currentTimeMillis(); // 如果当前时间小于上次生成ID的时间戳，说明系统时钟回拨，抛出异常 if (timestamp < lastTimestamp) { throw new RuntimeException("Clock is moving backwards. Rejecting requests until " + lastTimestamp); } // 如果在同一毫秒内，增加序列号 if (lastTimestamp == timestamp) { sequence = (sequence + 1) & SEQUENCE_MASK; // 使用位运算确保序列号循环 // 如果序列号溢出，等待下一毫秒 if (sequence == 0) { timestamp = waitForNextMillis(lastTimestamp); } } else { sequence = 0L; // 如果是新的毫秒，重置序列号 } lastTimestamp = timestamp; // 更新上次生成ID的时间戳 // 组合ID return ((timestamp - EPOCH) << (MACHINE_ID_BITS + SEQUENCE_BITS)) | (machineId << SEQUENCE_BITS) | sequence; } private long waitForNextMillis(long lastTimestamp) { long timestamp = System.currentTimeMillis(); while (timestamp <= lastTimestamp) { timestamp = System.currentTimeMillis(); // 等待下一毫秒 } return timestamp; } }
2，雪花算法变种
【雪花算法-修正版】
/* * Long: 1,000,000,000,000,000,000-9,223,372,036,854,775,807，共19位 * 雪花算法变种实现: {HHmmssSSS-9位}+{系统号-2位}+{毫秒递增seq-4位}+{随机数-3位} * */ public class TiDBRandomPrimaryKeyGenerator { //机器号 private String machineId; public static AtomicInteger machineIndex = new AtomicInteger(0); //序列号 private long sequence = 0L; //上一个时间戳，用于保证同一毫秒内序列号不重复。 private long lastTimestamp = -1L; //序列号最大值 private static final long SEQUENCE_MASK = 9999; private String pattern = "HHmmssSSS"; public TiDBRandomPrimaryKeyGenerator() { machineId = getInerMachineId(); this.sequence = (long) (Math.random() * SEQUENCE_MASK); } public synchronized long nextId() { Date currentDate = new Date(); String dateTime = DateFormatUtils.format(currentDate, pattern); String seq = getInerSequence(currentDate.getTime()); String random = generateRandomString(3); StringBuilder id = new StringBuilder(32); id.append(dateTime).append(machineId).append(seq).append(random); return Long.parseLong(String.valueOf(id)); } private synchronized String getInerSequence(long timestamp) { if (timestamp < lastTimestamp) { throw new RuntimeException("Clock moved backwards. Refusing to generate id for " + (lastTimestamp - timestamp) + " milliseconds"); } // 如果是同一毫秒内的时间戳 if (lastTimestamp == timestamp) { sequence = (sequence + 1) % SEQUENCE_MASK; } else { sequence = 0; } lastTimestamp = timestamp; return String.format("%04d", sequence); } private String getInerMachineId( ) { String machineId = String.format("%02d", machineIndex.intValue()); this.machineIndex.incrementAndGet(); return machineId; } public static String generateRandomString(int length) { Random random = new Random(); StringBuilder sb = new StringBuilder(); for (int i = 0; i < length; i++) { // 生成0到9之间的随机数，并转换为字符 char randomChar = (char) (random.nextInt(10) + '0'); sb.append(randomChar); } return sb.toString(); } }
3，基于DB自增
算法：{机器号}+{数据库id自增}+{伪随机}
@Service public class UniqueIDGenerator { @Autowired PrefixRepository prefixRepository; // 队列最小容量阈值 private static final int QUEUE_MIN_SIZE = 10000; private static final int QUEUE_CAPACITY = 12000; private static final int RANDOM_UPPER_LIMIT = 9999; // 递增前缀 private long prefix; // 随机数队列 private BlockingQueue<String> idQueue = new ArrayBlockingQueue<>(QUEUE_CAPACITY); // 线程安全锁 private final Lock lock = new ReentrantLock(); public String getUniqueId() throws InterruptedException { String uniqueId; uniqueId = idQueue.poll(); if (uniqueId == null) { lock.lock(); try { refillQueue(); } finally { lock.unlock(); } uniqueId = idQueue.take(); } return uniqueId; } private long fetchDBPrefix() { Prefix prefix_from_db = prefixRepository.findById(1).get(); this.prefix = prefix_from_db.getNext_value(); prefix_from_db.setNext_value(prefix_from_db.getNext_value() + 1); prefixRepository.save(prefix_from_db); return prefix; } private void refillQueue() { //前1w个已经用完，重新获取前缀 prefix = fetchDBPrefix(); Random random = new Random(); while (idQueue.size() < QUEUE_MIN_SIZE) { int randomPart = random.nextInt(RANDOM_UPPER_LIMIT + 1); String id = String.format("%014d", prefix) + String.format("%04d", randomPart); idQueue.offer(id); } } }public interface PrefixRepository extends JpaRepository<Prefix, Integer> {}@Entity @Table(name = "unique_id_table") public class Prefix{ @Id @Column(name = "id") @GeneratedValue(strategy = GenerationType.IDENTITY) private int id; @Column(name = "next_value") private long next_value; public int getId() { return id; } public void setId(int id) { this.id = id; } public long getNext_value() { return next_value; } public void setNext_value(long nextValue) { this.next_value = nextValue; } }spring.datasource.driver-class-name=org.h2.Driver spring.datasource.url=jdbc:h2:mem:test spring.datasource.username=sa spring.datasource.password= spring.sql.init.mode=always spring.sql.init.schema-locations=classpath:/sql/schema.sql spring.sql.init.data-locations=classpath:/sql/data.sql spring.jpa.show-sql=true spring.jpa.hibernate.ddl-auto=updateDROP TABLE IF EXISTS `unique_id_table`; CREATE TABLE `unique_id_table` ( `id` int, `next_value` bigint ); INSERT INTO `unique_id_table` (`id`, `next_value`) VALUES (1,1);