一致性hash算法java实现

思路：利用JDK中TreeMap的排序功能进行hash环的映射

实现代码如下：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94

packagecom.sccoder.algorithm;   importjava.util.SortedMap; importjava.util.TreeMap;   importorg.apache.commons.codec.digest.DigestUtils;   /**  *  * @author 亦凡 一致性hash算法的实现  */ publicclass ConsistencyHash<T> {       /**      * 存储服务器信息，key为服务器的hash值，TreeMap根据key进行了排序，便于查找映射节点      */     privateTreeMap<Long, T> serverNodes = null;       publicConsistencyHash() {         serverNodes = newTreeMap<>();     }       publicConsistencyHash(T[] servers) {         serverNodes = newTreeMap<>();         for(T server : servers) {             addServer(server);         }     }       /**      * 映射服务器到hash环上      */     publicvoid addServer(T server) {         serverNodes.put(hash(server), server);     }       /**      * 从hash环上移除服务器      */     publicvoid removeServer(T server) {         serverNodes.remove(hash(server));     }       /**      * 根据key的hash值在hash环上的映射查找key映射的服务器      *      * @param keyHash      * @return      */     publicT getServerNode(Long keyHash) {         if(serverNodes == null) {             returnnull;         }         SortedMap<Long, T> tailMap = serverNodes.tailMap(keyHash);         if(tailMap.isEmpty()) {             keyHash = serverNodes.firstKey();         }else{             keyHash = tailMap.firstKey();         }         returnserverNodes.get(keyHash);     }       /**      * 打印server节点映射顺序      */     publicvoid printServerMapOrder() {         System.out.println(serverNodes);     }       /**      * 计算hash      *      * @param obj      * @return      */     publicstatic long hash(Object obj) {         byte[] data = DigestUtils.md5(obj.toString().getBytes());         returndata[0] | ((long) data[1] << 8) | ((long) data[2] << 16)                 | ((long) data[3] << 24) | ((long) data[4] << 32)                 | ((long) data[5] << 40) | ((long) data[6] << 48)                 | ((long) data[7] << 56);     }       /**      * byte数组转化为long数组      *      * @param byteArray      * @return      */     publicstatic long byteToLong(byte[] byteArray) {         returnLong.parseLong(newString(byteArray));     }   }

下面用数据做一下测试:

加入5条服务器信息，并将5条数据映射到服务器上，查看映射情况。

然后移除server3，查看数据映射情况变化。

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

packagecom.sccoder.algorithm;   /**  *  * @author 亦凡  *  */ publicclass Test {       publicstatic void main(String[] args) {           String[] servers = newString[] { "Server 1:192.168.1.1",                 "Server 2:192.168.1.2","Server 3:192.168.1.3",                 "Server 4:192.168.1.4","Server 5:192.168.1.5" };           ConsistencyHash<String> consHash = newConsistencyHash<>(servers);         System.out.println("服务器映射信息：");         consHash.printServerMapOrder();         System.out.println("数据映射信息：");         showDataMap(consHash);         // 移除server2         consHash.removeServer(servers[2]);         System.out.println("移除server 3后数据映射信息：");         showDataMap(consHash);       }       publicstatic void showDataMap(ConsistencyHash<String> consHash) {         for(inti = 0; i < 5; i++) {             System.out.println("Data"+ i + " mapped at "                     + consHash.getServerNode(ConsistencyHash.hash("Data"+ i)));         }       }   }

打印结果如下：

服务器映射信息：
{-1886886=Server 4:192.168.1.4, -27899=Server 2:192.168.1.2, -81=Server 5:192.168.1.5, -46=Server 1:192.168.1.1, -19=Server 3:192.168.1.3}
数据映射信息：
Data0 mapped at Server 5:192.168.1.5
Data1 mapped at Server 5:192.168.1.5
Data2 mapped at Server 3:192.168.1.3
Data3 mapped at Server 4:192.168.1.4
Data4 mapped at Server 4:192.168.1.4
移除server 3后数据映射信息：
Data0 mapped at Server 5:192.168.1.5
Data1 mapped at Server 5:192.168.1.5
Data2 mapped at Server 4:192.168.1.4
Data3 mapped at Server 4:192.168.1.4
Data4 mapped at Server 4:192.168.1.4

       server3在hash环上的顺序在server4前面，所以移除server3后，映射到server3上面的数据就会映射到server4上面。在服务器崩溃之后，只会影响这一台服务器上面数据的可用性。其他服务器是不会受影响的。