How to use Redis

Usage scenarios

In my project, there is a function provided for Autocomplete, and the amount of data is probably tens of thousands. In this article, I use the example of name retrieval to illustrate. For the list, please click on the Demo from the author of Redis.

Such a list is full of user names. For example, there is a user object in our system:

public Class User
{     public string Id{get; set;}  
     public string Name {get; set;}
     ....     public string UserHead {get; set;}    
}

The system needs a drop-down list of users, which cannot be done due to the large amount of data. It is displayed once, so an AutoComplete function is added. The cache can be stored directly in local memory. There is no need to use a centralized cache like Redis, so the cache structure will be simpler.

var users = new List<User>{...};//读到一个用户列表MemoryCache.Set("capqueen:users", users);//放入内存//读取var users = MemoryCache.Get<List<User>>("capqueen:users");

Because they are all in memory, it is enough to store the List directly. When searching It can also be directly as follows:

var findUsers = users.Where(user => user.Name.StartWith("A")).ToList();例如输入的字符是 “A“

It is quite simple. There is no need to consider how to store and the stored data structure. However, once we move to a centralized caching service like Redis, we need to rethink how we store it.

Option 1: Similar memory-based cache implementation.

The Redis link library used in this article is StactkExchange.Redis, an open source product from StackOverFlow.

var db = redis.GetDataBase();//获取0数据库var usersJson = JsonConvert.SerializeObject(users)//序列化db.StringSet("capqueen:users", usersJson);//存储var usersString = db.StringGet("capqueen:users");
var userList = JsonConvert.DeserializeObject<List<User>>(users);//反序列化

There is no logical problem with the above method, and the compilation can pass. But if you think about it carefully, Redis is an independent cache service and is separated from appSever. This kind of reading method is a burden on the IO of the redis server, and even such reading is much slower than the local memory cache.

How to solve it? Just imagine that the essence of key-value lies in Key, so for List, items should be stored separately.

Option 2: Keys fuzzy matching.

After browsing the Redis command documentation (see Reference 4), he was surprised to find the Keys command, which made him immediately modify his plan. First, we need to establish the keyword to be searched as a key. Here I define the key as "capqueen:user:{id}:{name}", where the items in {} need to be replaced with the corresponding attributes of the item. The code is as follows:

var redis = ConnectionMultiplexer.Connect("localhost");var db = redis.GetDatabase();           
var users = new List<User> { new User{Id = 6, Name = "aaren", Age=10},    new User{Id = 7, Name = "issy", Age=11},    new User{Id = 8, Name = "janina", Age=13},    new User{Id = 9, Name = "karena", Age=14}
};

users.ForEach(item => { 
   var key = string.Format("capqueen:user:{0}:{1}", item.Id, item.Name);   var value = JsonConvert.SerializeObject(item);
   db.StringSet(key, value);
});

All users are stored in separate Key-Value methods, so how to use Keys to search? Let’s take a look at the Keys command of Redis:

KEYS pattern

查找所有符合给定模式 pattern 的 key 。

KEYS * 匹配数据库中所有 key 。
KEYS h?llo 匹配 hello ， hallo 和 hxllo 等。
KEYS h*llo 匹配 hllo 和 heeeeello 等。
KEYS h[ae]llo 匹配 hello 和 hallo ，但不匹配 hillo 。
特殊符号用 \ 隔开

In other words, Keys can perform simple fuzzy matching, so our search here can be changed to the following method:

var redis = ConnectionMultiplexer.Connect("192.168.10.178");var db = redis.GetDatabase();var server = redis.GetServer("192.168.10.178", 6379);var keys = server.Keys(pattern: "capqueen:user:*:a*");var values = db.StringGet(keys.ToArray());//反序列化var jsonValues = new StringBuilder("[");
values.ToList().ForEach(item => jsonValues.Append(item).Append(","));
jsonValues.Append("]");var userList = JsonConvert.DeserializeObject<List<User>>(jsonValues.ToString());

Note the above In the code, because each value is a json, in order to increase the efficiency of conversion, I first process it into json arry and then deserialize it.

This solution indeed solved my current problem. However, I noticed a passage in the Redis document:

KEYS is very fast, but when used in a large database It may still cause performance problems. If you need to find a specific key from a data set, you are better off using Redis's set structure (set) instead.

The above is the detailed content of How to use Redis. For more information, please follow other related articles on the PHP Chinese website!