一致性hash算法
介绍
http://blog.csdn.net/cywosp/article/details/23397179/
http://www.cnblogs.com/haippy/archive/2011/12/10/2282943.html
http://blog.csdn.net/caigen1988/article/details/7708806
http://www.blogjava.net/hello-yun/archive/2012/10/10/389289.html
java 实现
http://www.iteye.com/topic/1132274
http://blog.csdn.net/wuhuan_wp/article/details/7010071
http://www.oschina.net/code/snippet_730640_22941
思路
1 定义一个Hash算法
2 定义一个Map,以IP等作为key,hash出N个key复本作为Key(value为真实的节点)存入Map
for (int i = 0; i < numberOfReplicas; i++) {
circle.put(hashFunc.hash(node.toString() + i), node);
}
3取值:hash传入的值,Map.get(hash(key)),如果没有,取tailMap的第一个key,再Map.get(hash(key))
public T get(Object key) {
readReadWriteLock.readLock().lock();
try{
if (circle.isEmpty()) {
return null;
}
int hash = hashFunc.hash(key);
if (!circle.containsKey(hash)) {
SortedMap<Integer, T> tailMap = circle.tailMap(hash); // 返回此映射的部分视图,其键大于等于
hash = tailMap.isEmpty() ? circle.firstKey() : tailMap.firstKey();
}
// 正好命中
return circle.get(hash);
}finally{
readReadWriteLock.readLock().unlock();
}
}