Redis (replicate dictionary server) cache

Redis（REmote DIctionary Server, Remote data services ）

        Redis（REmote DIctionary Server, Remote data services ） Yes, it is C Language Developed an open source high-performance   Key value pair （key-value） database ,Redis Save all data in memory . Official test data ,50 Individual concurrent execution 100000 A request , The speed of reading is zero 110000 Time /s, The speed of writing is 81000 Time /s , And Redis Multiple key value data types are provided to meet the storage requirements in different scenarios , So far, Redis The supported key value data types are as follows ：key always String type ,redis Supported data types Refers to value Types of data that can be stored .

Redis The default is 16 A database , The number is from 0~15. 【 The default is 0 Signal library 】

• keys *: Query all key

• exists key: Judge whether there is a specified key If there is a return 1, Otherwise return to 0

• expire key Number of seconds : Set this key Lifetime in cache

• ttl key: Show assignments key The rest of the time
  If the return value is Positive numbers ： Appoint key The rest of the time
  If the return value is -1： Never expire

  If the return value is -2： Does not exist or has expired

• del key: Delete the specified key

• rename key new key: rename

• type key: Judge a key The type of

• ping : Test whether the connection is connected

• select index: Switch Library

• move key index: hold key Move to library number (index Is the library number )

• flushdb: Clear the current database

• flushall: Clear all databases under the current instance

Redis data type

String type	string
List the type	List	similar LinkedList
Hash type / Hash type	hash	similar HashMap, Suitable for storing objects ( attribute , Property value )
Collection types （ disorder ）	set	similar HashSet, Disorder and uniqueness
Ordered set type	zset/sorted_set	Orderly and unique

Redis The persistence of

        Redis Because it stores all the data in memory , In order to make Redis After the restart, the data can still be guaranteed not to be lost , Data need to be Sync from memory to hard disk ( file ) in , The process is persistence .

Redis Support two ways of persistence , One is RDB The way , One is AOF The way .

RDB Persistence ： This mechanism can generate point in time snapshot of data set in specified time interval （point-in-time snapshot）.

AOF Persistence ： Record all write commands executed by the server , And when the server starts , Restore the dataset by reexecuting these commands .

RDB Persistence mechanism : Default It's on , No configuration required

keyword	Time ( second )	key Modify quantity	explain
save	900	1	Every time 900 second (15 minute ) There are at least 1 individual key change , be dump memory dump
save	300	10	Every time 300 second (5 minute ) There are at least 10 individual key change , be dump memory dump
save	60	10000	Every time 60 second (1 minute ) There are at least 10000 individual key change , be dump memory dump

AOF Persistence mechanism ： To use, you need to configure , Open manually AOF Persistence

keyword	Time to persist	explain
appendfsync	always	Every time the update command is executed , Persistent once
appendfsync	everysec	Persist once per second
appendfsync	no	Non persistence

RDB Persistence and AOF Persistence is different

	advantage	shortcoming
RDB	1.RDB Speed ratio when recovering large data sets AOF It's faster to recover ( Because its files are better than AOF Small ) 2.RDB It's better than AOF Better	1.RDB The persistence of is not timely ( At regular intervals ), There may be data loss 2.RDB During persistence, if the file is too large, it may cause server blocking , Stop receiving client requests
AOF	1.AOF The durability is more durable ( Every second. or Save once per operation ) 2.AOF The contents of the document are very easy to read , Analyze the document （parse） It's easy too ,AOF It's an incremental operation	1.AOF The volume of the file is usually larger than RDB Volume of file 2.AOF May be slower than RDB

RDB and AOF The difference between ？
RDB： Default on , Perform persistence operation within a certain time interval , Synchronize the data in memory into a binary file , The file size is small , Faster recovery , Good performance , But it is prone to data loss .
AOF： Manual opening required , Persistence can be achieved per second or after each operation , The default is to persist once per second , take redis The operation command of is saved to an incremental file , File size is large , Slow recovery , Low performance , But the database integrity is good .

Jedis Is the use of Java operation Redis The client of （ tool kit :Jedis=JDBC+ drive ）

1.Jedis To get started

demand : Use java Code operation Redis Add ( Change ) Delete query  

step :

        1. Create project , Import jedis Of jar package
        2. establish jedis object  new Jedis(String host,int port);
        3. Calling method
        4. Close the object Release resources

public class Demo01 {
    public static void main(String[] args) {

        //1. establish jedis object 
        Jedis jedis = new Jedis("localhost",6379);
        //2. Calling method 
        //2.1：string Type operation 
        // obtain value
        String akey = jedis.get("akey");
        System.out.println("akey = " + akey);
        // Judge key Whether there is 
        Boolean flag = jedis.exists("akey");
        System.out.println("akey Whether there is ："+flag);
        //jedis.del("akey");   Delete key
        // Storage key-value
        jedis.set("dkey","ddd");

        //2.2：hash type 
        jedis.hset("user1","name","zs");
        String name = jedis.hget("user1", "name");
        System.out.println("name = " + name);

        //2.3:list type 
        jedis.rpush("list1","a","b","c");
        System.out.println(jedis.lindex("list1", 2));

        //2.4:set type 
        jedis.sadd("set1","a","b","c");
        Set<String> set1 = jedis.smembers("set1");
        for (String s : set1) {
            System.out.println(s);
        }

        //3. Close the object 
        jedis.close();
    }
}

2.Jedis The use of connection pools -Jedis Advanced

demand : from Jedis From the connection pool jedis

step :

1. establish JedisPoolConfig object

2. establish JedisPool object

3. obtain jedis object

4. Calling method

5. The return jedis Object to Jedis Connection pool

public class Demo02 {
    public static void main(String[] args) {
        //1.  establish JedisPoolConfig object 
        JedisPoolConfig config = new JedisPoolConfig();
        config.setMaxTotal(5);
        config.setMaxWaitMillis(5000);
        //2.  establish JedisPool object 
        JedisPool jedisPool = new JedisPool(config, "localhost", 6379);
        //3.  obtain jedis object 
        Jedis jedis = jedisPool.getResource();
        //4.  Calling method    demand ： If akey There is   obtain akey Value   ;  If akey non-existent   Create a akey=aaa Deposit in redis in 
        if(jedis.exists("akey")){
            System.out.println(jedis.get("akey"));
        }else{
            jedis.set("akey","aaa");
        }
        //5.  The return jedis Object to Jedis Connection pool 
        jedis.close();
    }
}

3.JedisUtils Tool class extraction -Jedis Advanced

Purpose : 1. Make sure there is only one pool 2. get jedis object 3. The return

step :

1. establish JedisUtils Tool class

2. obtain jedis object

3. The return jedis object

4. Make sure there is only one pool Use static code blocks to create connection pool objects

import redis.clients.jedis.Jedis;
import redis.clients.jedis.JedisPool;
import redis.clients.jedis.JedisPoolConfig;

/**
 * Jedis Tool class 
 *  1. Provide access to jedis Object method   getJedis();
 *  2. Provide return jedis Object method   close(Jedis jedis);
 *   Optimize ： Give Way jedisPool Object has only one    Use the singleton mode    Declare a static variable JedisPool object    Then initialize with static code blocks JedisPool object 
 */
public class JedisUtils {

    private static JedisPool jedisPool;

    static{
        //1.  establish JedisPoolConfig object 
        JedisPoolConfig config = new JedisPoolConfig();
        config.setMaxTotal(5);
        config.setMaxWaitMillis(3000);
        //2.  establish JedisPool object 
        jedisPool = new JedisPool(config,"localhost",6379);
    }

    /**
     *  obtain jedis object 
     * @return
     */
    public static Jedis getJedis(){
        //3.  obtain jedis object 
        return jedisPool.getResource();
    }

    /**
     *  The return jedis object 
     * @param jedis
     */
    public static void close(Jedis jedis){
        if(jedis!=null){
            jedis.close();
        }
    }
}

Redis The pattern of

• Standaloan Pattern （ standalone mode ） Stand alone mode refers to running in a single server Redis Program , Is the most primitive and basic mode . All services are connected to one Redis service , This mode is not applicable to production . In case of downtime , Memory explosion , It may cause all connections to change redis The cache failure of the service caused an avalanche

• Master-slave Pattern （ A master-slave mode ） Can achieve read-write separation , The data backup . But it's not 「 High availability 」 Of

• Sentinel Pattern （ Sentinel mode ） It can be regarded as master-slave mode 「 High availability 」 edition , It introduces Sentinel To the whole Redis Service cluster for monitoring . But because there is only one master node , Therefore, there is still a write bottleneck .

• Cluster Pattern （ Cluster pattern ） It not only provides a highly available means , At the same time, the data is saved in pieces in each node , It can support high concurrency write and read . Of course, the implementation is also the most complex .

1.Master-slave Pattern （ A master-slave mode ）

Redis The master-slave mode of refers to master-slave replication :

1. From the server to the master server SYNC  or  PSYNC  command
2. Primary server execution  BGSAVE command , Generate RDB file , And use the cache to record all write commands from now on
3. RDB After the file is generated , The master server will send it to the slave server
4. Load... From the server RDB file , Update your database status synchronously to the master server  BGSAVE The state of being ordered .
5. The master server sends all write commands of the buffer to the slave server , The slave service will execute these write commands , The database state is synchronized with the latest state of the main server .

SYNC And PSYNC The difference between

• SYNC
  Re initiate from the server to the master server  SYNC command , The main server will All the data Regenerate again RDB The snapshot is sent to the server for synchronization
• PSYNC
  Re initiate from the server to the master server  PSYNC command . master server Judge whether it is necessary to complete resynchronization or only send the write command executed during the disconnection to the slave server according to the deviation of the data of both parties .

PSYNC How to realize partial resynchronization ？

         The realization of partial resynchronization mainly depends on three parts ：

                1. Record the copy offset

                  2. Copy backlog buffer

                3. Recording server ID

2.Sentinel Pattern （ Sentinel mode ）

3.Cluster Pattern （ Cluster pattern ） 

        Redis Cluster pattern of In order to solve the horizontal expansion of the system and the storage of massive data , If you have a lot of data , So you can use it redis cluster.