ShardedJedisPool详解

ShardedJedisPool详解

ShardedJedisPool类有四个构造函数

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
public ShardedJedisPool(final GenericObjectPoolConfig poolConfig, List<JedisShardInfo> shards) {
this(poolConfig, shards, Hashing.MURMUR_HASH);
}

public ShardedJedisPool(final GenericObjectPoolConfig poolConfig, List<JedisShardInfo> shards,
Hashing algo) {
this(poolConfig, shards, algo, null);
}

public ShardedJedisPool(final GenericObjectPoolConfig poolConfig, List<JedisShardInfo> shards,
Pattern keyTagPattern) {
this(poolConfig, shards, Hashing.MURMUR_HASH, keyTagPattern);
}

public ShardedJedisPool(final GenericObjectPoolConfig poolConfig, List<JedisShardInfo> shards,
Hashing algo, Pattern keyTagPattern) {
super(poolConfig, new ShardedJedisFactory(shards, algo, keyTagPattern));
}

其实,最后都是调用父类的构造函数

1
super(poolConfig, new ShardedJedisFactory(shards, algo, keyTagPattern));

输入参数说明:

1
2
3
4
5
6
7
final GenericObjectPoolConfig poolConfig:pool的一些设置(maxTotal、maxIdle、minIdle)
List<JedisShardInfo> shards:为Redis实例的IP、端口以及超时时间等信息
Hashing algo:根据Key计算hash值,从而决定具体存入那个Redis实例
Pattern keyTagPattern:Key标签模式,默认是取{}中内容去计算对应Hash值,
当key='{key1_0}_0_0'时,此时将根据’key1_0'来计算对应hash值,而不是根据{key1_0}_0_0来计算hash值,
从而决定将存入到哪个Redis实例中,如果需要将数据进行排序,将要利用这个构造函数,因为只有将需要排序
的数据存入同一个Redis实例中,排序才是准确的。