Redis从入门到放弃系列(二) Hash

本文例子基于：5.0.4 Hash是Redis中一种比较常见的数据结构,其实现为hashtable/ziplist,默认创建时为ziplist,当到达一定量级时,redis会将ziplist转化为hashtable

Redis从入门到放弃系列(一) String

首先让我们来看一下该如何在redis里面使用Hash类型

//将hash表中key的域field的值设为value
//如果key不存在，一个新的哈希表被创建并进行HSET操作
//如果域field已经存在于哈希表中，旧值将被覆盖
hset key field value

代码示例：

//创建不存在的field
>hset user:1 id 1
(integer) 1
//覆盖原先的field
>hset user:1 id 2
(integer) 0
>hget user:1 id
"2"
//获取不存在的field
>hget user:1 not_exist
(nil)
----------------------------------
// hsetnx key field value
//当不存在该field 设置成功返回1 ，否则返回0
> hsetnx user:1 id 1
(integer) 1
> hsetnx user:1 id 1
(integer) 0
> hget user:1 id
"1"
----------------------------------
// hmset key field value [field value ....]
//批量设置多个键值对
>HMSET user:1 id 1 name "黑搜丶D" wechat "black-search"
OK
----------------------------------
//hget key field
//获取hash表key中给定的field的值
>hget user:1 id 
"1"
----------------------------------
// hmget key field[field...]
//按照我们输入的field的顺序返回
>hmget user:1 name wechat id  not_exist
1) "黑搜丶D"
2) "black-search"
3) "1"
4) (nil)
----------------------------------
// hdel key field 删除返回被成功移除的域的数量 
> hgetall user:1
1) "id"
2) "1"
3) "name"
4) "black-search"
> HDEL user:1 name
(integer) 1
> HDEL user:1 name
(integer) 0
----------------------------------
// HINCRBY key field increment
// 为hash表某个整数类型的field增加increment ,返回增加increment之后的大小
> hset user:1 wechat "black-search"
(integer) 1
> HINCRBY user:1 wechat 2
(error) ERR hash value is not an integer
> HINCRBY user:1 id 21
(integer) 22
> hget user:1 id
"22"

至此,redis hash的用法先告一段落.

debug object key

本文开头的时候讲默认创建为ziplist,当达到一定的量级转化为hashtable,那么具体是在什么时候才会转化成hashtable呢？

# Hashes are encoded using a memory efficient data structure when they have a
# small number of entries, and the biggest entry does not exceed a given
# threshold. These thresholds can be configured using the following directives.
hash-max-ziplist-entries 512
hash-max-ziplist-value 64

从上文我们可以知道，只有当我们满足以下两个条件会将ziplist转化为hashtable结构

1. 保存的所有键值对个数小于 512个 (这个限制是由 hash-max-ziplist-entries 参数控制，默认 512)
2. 保存的所有键值对的长度都小于 64 字节（这个限制是由 hash-max-ziplist-value 参数控制，默认 64）

// 这里测试当键值对小于等于512时，hash的类型
@RequestMapping("/")
public void test(){
	List<Long> list = redisTemplate.executePipelined(new RedisCallback<Long>() {
		@Override
		public Long doInRedis(RedisConnection redisConnection) throws DataAccessException {
			redisConnection.openPipeline();
			for (int i=0;i<512;i++){
				redisConnection.hSet("key".getBytes(),("field"+i).getBytes(),"value".getBytes());
			}
			return null;
		}
	});
	System.out.println("结束");
}
//我们发现这里hash的类型就是ziplist
> debug object key
Value at:0xbc6f80 refcount:1 encoding:ziplist serializedlength:2603 lru:14344435 lru_seconds_idle:17
//让我们调大一下循环的次数，改为513，我们发现
> debug object key
Value at:0xbc6f80 refcount:1 encoding:hashtable serializedlength:7587 lru:14344656 lru_seconds_idle:4

源码解析

//首先我们来看一下dict的结构
typedef struct dict {
    dictType *type;
    void *privdata;
    dictht ht[2];
    long rehashidx; /* rehashing not in progress if rehashidx == -1 */
    unsigned long iterators; /* number of iterators currently running */
} dict;
typedef struct dictType {
    uint64_t (*hashFunction)(const void *key);
    void *(*keyDup)(void *privdata, const void *key);
    void *(*valDup)(void *privdata, const void *obj);
    int (*keyCompare)(void *privdata, const void *key1, const void *key2);
    void (*keyDestructor)(void *privdata, void *key);
    void (*valDestructor)(void *privdata, void *obj);
} dictType;
/* This is our hash table structure. Every dictionary has two of this as we
 * implement incremental rehashing, for the old to the new table. */
typedef struct dictht {
    dictEntry **table;
    unsigned long size;
    unsigned long sizemask;
    unsigned long used;
} dictht;
typedef struct dictEntry {
    void *key;
    union {
        void *val;
        uint64_t u64;
        int64_t s64;
        double d;
    } v;
    struct dictEntry *next;
} dictEntry;

从以上我们可以知道，dict里面包含了两个dictht(ps:hashtable),通常情况下只有一个dictht有值.但是当dict扩容/缩容的时候,需要分配新的dictht，然后渐进式搬迁，当迁移结束之后，旧的dictht被删除，只保留新的dictht dict如何解决hash冲突呢？其实原理跟Java的HashMap是一样的，采用数组+链表的方式去解决

渐进式rehash

我们知道，redis是单进程的,如果要将一个大的字典扩容是会比较耗时的,那么有可能就会将其他请求挂起。所以redis采用渐进式rehash来完成这一项艰巨任务~

dictEntry *dictAddRaw(dict *d, void *key, dictEntry **existing)
{
    long index;
    dictEntry *entry;
    dictht *ht;
    //这里每次都会进行搬迁~
    if (dictIsRehashing(d)) _dictRehashStep(d);

    /* Get the index of the new element, or -1 if
     * the element already exists. */
    if ((index = _dictKeyIndex(d, key, dictHashKey(d,key), existing)) == -1)
        return NULL;

    /* Allocate the memory and store the new entry.
     * Insert the element in top, with the assumption that in a database
     * system it is more likely that recently added entries are accessed
     * more frequently. */
    //当字典处于搬迁中，将新添加的元素挂到新的数组下面
    ht = dictIsRehashing(d) ? &d->ht[1] : &d->ht[0];
    entry = zmalloc(sizeof(*entry));
    entry->next = ht->table[index];
    ht->table[index] = entry;
    ht->used++;

    /* Set the hash entry fields. */
    dictSetKey(d, entry, key);
    return entry;
}

这样,在客户端每次请求(hset/hdel等)都会去判断是否需要搬迁,那么当客户端不请求我们的时候,有可能没有完整的搬迁？no no no redis会在定时任务里面扫描处于rehash的dict，然后完成剩余的搬迁~代码如下

/* This function handles 'background' operations we are required to do
 * incrementally in Redis databases, such as active key expiring, resizing,
 * rehashing. */
void databasesCron(void) {
    /* Expire keys by random sampling. Not required for slaves
     * as master will synthesize DELs for us. */
    if (server.active_expire_enabled) {
        if (server.masterhost == NULL) {
            activeExpireCycle(ACTIVE_EXPIRE_CYCLE_SLOW);
        } else {
            expireSlaveKeys();
        }
    }

    /* Defrag keys gradually. */
    if (server.active_defrag_enabled)
        activeDefragCycle();

    /* Perform hash tables rehashing if needed, but only if there are no
     * other processes saving the DB on disk. Otherwise rehashing is bad
     * as will cause a lot of copy-on-write of memory pages. */
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
        /* We use global counters so if we stop the computation at a given
         * DB we'll be able to start from the successive in the next
         * cron loop iteration. */
        static unsigned int resize_db = 0;
        static unsigned int rehash_db = 0;
        int dbs_per_call = CRON_DBS_PER_CALL;
        int j;

        /* Don't test more DBs than we have. */
        if (dbs_per_call > server.dbnum) dbs_per_call = server.dbnum;

        /* Resize */
        for (j = 0; j < dbs_per_call; j++) {
            tryResizeHashTables(resize_db % server.dbnum);
            resize_db++;
        }

        /* Rehash */
        //重点在这里rehash
        if (server.activerehashing) {
            for (j = 0; j < dbs_per_call; j++) {
                int work_done = incrementallyRehash(rehash_db);
                if (work_done) {
                    /* If the function did some work, stop here, we'll do
                     * more at the next cron loop. */
                    break;
                } else {
                    /* If this db didn't need rehash, we'll try the next one. */
                    rehash_db++;
                    rehash_db %= server.dbnum;
                }
            }
        }
    }
}

应用场景

储存业务数据,我们发现其实hset的用法很简单,回顾上一讲最后的应用场景

//上一讲使用string 
>set user:1 '{"id":1,"name":"黑搜丶D","wechat":"black-search"}'
//让我们使用hash来实现相似的做法
> HMSET user:1 id 1 name "黑搜丶D" wechat "black-search"
OK
//获取key的某个field的值
>hget user:1 wechat
"black-search"
//获取到key的所有 field:value组合
> HGETALL user:1
1) "id"
2) "1"
3) "name"
4) "\xe9\xbb\x91\xe6\x90\x9c\xe4\xb8\xb6D"
5) "wechat"
6) "black-search"

相对于string的用法,我们使用hash get某个field或者set某个field会省很多带宽~