Twitter's distributed self-incrementing ID algorithm snowflake

Comparison of advantages and disadvantages of various primary keys

1. Database auto-incrementing primary key

Advantages: Simple; Unique; Incremental; Fixed Increment

Disadvantages: write performance determines the upper limit of the number of generated per second, poor expansion; distributed database, the master node hangs, there may be problems on the standby node (the master node writes successfully, the log is not synchronized to the standby node, resulting in duplicate ids)

Remarks: One write library can become multiple libraries to write at the same time. For example, three libraries of 1, 2, and 3 are written at the same time. The initial ids are 1, 2, and 3, respectively, and the self-increase degree is 3. This method ensures that the id is not repeated. However, the id does not increase absolutely, but increases in an overall trend; secondly, the pressure of writing is still very high, and mysql is easy to become a performance bottleneck.

 

2. The database generates IDs in batches

Advantages: high efficiency; reduce database pressure

Disadvantages: Security issues need to be considered to prevent duplicate ids from being obtained; if the business requirement is to generate only one id at a time, there is a performance problem

Remarks: Use the database to initialize a row of data, the initial value is 1, take 10 ids, add 10 to the value, and the caller takes the first 10 values ​​of the returned id value. The above is the idea of ​​batch generation id.

 

3.UUID

Pros: Local generation; high efficiency

Disadvantages: The UUID string is too long and has no practical significance; the increasing trend cannot be guaranteed; the established index query efficiency is low

 

4. Current time in milliseconds and microseconds

Advantages: local generation; low latency; high indexing performance

Disadvantage: After 1000 requests are made within 1 second, the id must be repeated, and the same is true for microseconds

 

5.zookeeper generates id

Using zookeeper to increase the version number is one of them. Create a node, incrementing the node version by 1 each time.

Advantages: Use zk clusters to solve single point problems

Disadvantages: low performance; id has an upper limit, 32-bit id is provided; zk service is required

Note: For the specific code, see  http://blog.csdn.net/gongzi2311/article/details/58144091

 

6.snowflake algorithm

Twitter open source distributed generation id algorithm.

Advantages: basically solve all problems

Disadvantages: The time of each node may be different, and the generated id is an overall trend of increasing

 

So I use the snowflake algorithm to generate the primary key, the code is as follows

The structure of snowflake is as follows (each part is separated by -):

0 - 0000000000 0000000000 0000000000 0000000000 0 - 00000 - 00000 - 000000000000

第一位为未使用，接下来的41位为毫秒级时间(41位的长度可以使用69年)，然后是5位datacenterId和5位workerId(10位的长度最多支持部署1024个节点） ，最后12位是毫秒内的计数（12位的计数顺序号支持每个节点每毫秒产生4096个ID序号）

一共加起来刚好64位，为一个Long型。(转换成字符串长度为18)

snowflake生成的ID整体上按照时间自增排序，并且整个分布式系统内不会产生ID碰撞（由datacenter和workerId作区分），并且效率较高。据说：snowflake每秒能够产生26万个ID。

/**
 * Twitter_Snowflake<br>
 * SnowFlake的结构如下(每部分用-分开):<br>
 * 0 - 0000000000 0000000000 0000000000 0000000000 0 - 00000 - 00000 - 000000000000 <br>
 * 1位标识，由于long基本类型在Java中是带符号的，最高位是符号位，正数是0，负数是1，所以id一般是正数，最高位是0<br>
 * 41位时间截(毫秒级)，注意，41位时间截不是存储当前时间的时间截，而是存储时间截的差值（当前时间截 - 开始时间截)
 * 得到的值），这里的的开始时间截，一般是我们的id生成器开始使用的时间，由我们程序来指定的（如下下面程序IdWorker类的startTime属性）。41位的时间截，可以使用69年，年T = (1L << 41) / (1000L * 60 * 60 * 24 * 365) = 69<br>
 * 10位的数据机器位，可以部署在1024个节点，包括5位datacenterId和5位workerId<br>
 * 12位序列，毫秒内的计数，12位的计数顺序号支持每个节点每毫秒(同一机器，同一时间截)产生4096个ID序号<br>
 * 加起来刚好64位，为一个Long型。<br>
 * SnowFlake的优点是，整体上按照时间自增排序，并且整个分布式系统内不会产生ID碰撞(由数据中心ID和机器ID作区分)，并且效率较高，经测试，SnowFlake每秒能够产生26万ID左右。
 */
public class SnowflakeIdWorker {

    // ==============================Fields===========================================
    /** 开始时间截 (2015-01-01) */
    private final long twepoch = 1420041600000L;

    /** 机器id所占的位数 */
    private final long workerIdBits = 5L;

    /** 数据标识id所占的位数 */
    private final long datacenterIdBits = 5L;

    /** 支持的最大机器id，结果是31 (这个移位算法可以很快的计算出几位二进制数所能表示的最大十进制数) */
    private final long maxWorkerId = -1L ^ (-1L << workerIdBits);

    /** 支持的最大数据标识id，结果是31 */
    private final long maxDatacenterId = -1L ^ (-1L << datacenterIdBits);

    /** 序列在id中占的位数 */
    private final long sequenceBits = 12L;

    /** 机器ID向左移12位 */
    private final long workerIdShift = sequenceBits;

    /** 数据标识id向左移17位(12+5) */
    private final long datacenterIdShift = sequenceBits + workerIdBits;

    /** 时间截向左移22位(5+5+12) */
    private final long timestampLeftShift = sequenceBits + workerIdBits + datacenterIdBits;

    /** 生成序列的掩码，这里为4095 (0b111111111111=0xfff=4095) */
    private final long sequenceMask = -1L ^ (-1L << sequenceBits);

    /** 工作机器ID(0~31) */
    private long workerId;

    /** 数据中心ID(0~31) */
    private long datacenterId;

    /** 毫秒内序列(0~4095) */
    private long sequence = 0L;

    /** 上次生成ID的时间截 */
    private long lastTimestamp = -1L;

    //==============================Constructors=====================================
    /**
     * 构造函数
     * @param workerId 工作ID (0~31)
     * @param datacenterId 数据中心ID (0~31)
     */
    public SnowflakeIdWorker(long workerId, long datacenterId) {
        if (workerId > maxWorkerId || workerId < 0) {
            throw new IllegalArgumentException(String.format("worker Id can't be greater than %d or less than 0", maxWorkerId));
        }
        if (datacenterId > maxDatacenterId || datacenterId < 0) {
            throw new IllegalArgumentException(String.format("datacenter Id can't be greater than %d or less than 0", maxDatacenterId));
        }
        this.workerId = workerId;
        this.datacenterId = datacenterId;
    }

    // ==============================Methods==========================================
    /**
     * 获得下一个ID (该方法是线程安全的)
     * @return SnowflakeId
     */
    public synchronized long nextId() {
        long timestamp = timeGen();

        //如果当前时间小于上一次ID生成的时间戳，说明系统时钟回退过这个时候应当抛出异常
        if (timestamp < lastTimestamp) {
            throw new RuntimeException(
                    String.format("Clock moved backwards.  Refusing to generate id for %d milliseconds", lastTimestamp - timestamp));
        }

        //如果是同一时间生成的，则进行毫秒内序列
        if (lastTimestamp == timestamp) {
            sequence = (sequence + 1) & sequenceMask;
            //毫秒内序列溢出
            if (sequence == 0) {
                //阻塞到下一个毫秒,获得新的时间戳
                timestamp = tilNextMillis(lastTimestamp);
            }
        }
        //时间戳改变，毫秒内序列重置
        else {
            sequence = 0L;
        }

        //上次生成ID的时间截
        lastTimestamp = timestamp;

        //移位并通过或运算拼到一起组成64位的ID
        return ((timestamp - twepoch) << timestampLeftShift) //
                | (datacenterId << datacenterIdShift) //
                | (workerId << workerIdShift) //
                | sequence;
    }

    /**
     * 阻塞到下一个毫秒，直到获得新的时间戳
     * @param lastTimestamp 上次生成ID的时间截
     * @return 当前时间戳
     */
    protected long tilNextMillis(long lastTimestamp) {
        long timestamp = timeGen();
        while (timestamp <= lastTimestamp) {
            timestamp = timeGen();
        }
        return timestamp;
    }

    /**
     * 返回以毫秒为单位的当前时间
     * @return 当前时间(毫秒)
     */
    protected long timeGen() {
        return System.currentTimeMillis();
    }

    //==============================Test=============================================
    /** 测试 */
    public static void main(String[] args) {
        SnowflakeIdWorker idWorker = new SnowflakeIdWorker(0, 0);
        for (int i = 0; i < 1000; i++) {
            long id = idWorker.nextId();
            System.out.println(Long.toBinaryString(id));
            System.out.println(id);
        }
    }
}