Redis Sentinel安裝與部署,實現redis的高可用

前言

  對於生產環境,高可用是避免不了要面對的問題,不管什麼環境、服務,只要用於生產,就須要知足高可用;此文針對的是redis的高可用。html

  接下來會有系列文章,該系列是對spring-session實現分佈式集羣session的共享的完整闡述,同時也引申出緩存的實現;而此篇是該系列的第一篇。java

  github地址:https://github.com/youzhibing/redislinux

環境準備

  redis版本:redis-3.0.0git

  linux:centos6.7github

  ip:192.168.11.202, 一臺服務器上搭建搭建所有redis實例,包括數據節點實例以及哨兵(sentinel)實例redis

  客戶端jedis,基於spring-bootspring

redis主從複製

  搭建一主二從的主從環境apache

  一、redis安裝

    安裝很簡單,網上資料不少,redis官網也有說明;主要就是3步:解壓,make,make installcentos

  二、數據節點配置文件

    redis解壓後,redis home目錄下有redis配置的樣例文件,咱們不直接在此文件上就行修改,在redis home目錄下新建文件夾master_slave,將配置文件都放於此目錄下緩存

    master配置文件:redis-6379.conf

port 6379
bind 192.168.11.202
requirepass "myredis"
daemonize yes
logfile "6379.log"
dbfilename "dump-6379.rdb"
dir "/opt/soft/redis/data"

#如若master設置了認證密碼,那麼全部redis數據節點都配置上masterauth屬性
masterauth "myredis"

    slave-1配置文件:redis-6380.conf

port 6380
bind 192.168.11.202
requirepass "myredis"
daemonize yes
logfile "6380.log"
dbfilename "dump-6380.rdb"
dir "/opt/soft/redis/data"

#如若master設置了認證密碼,那麼全部redis數據節點都配置上masterauth屬性
masterauth "myredis"
slaveof 192.168.11.202 6379

    slave-2配置文件:redis-6381.conf

port 6381
bind 192.168.11.202
requirepass "myredis"
daemonize yes
logfile "6381.log"
dbfilename "dump-6381.rdb"
dir "/opt/soft/redis/data"

#如若master設置了認證密碼,那麼全部redis數據節點都配置上masterauth屬性
masterauth "myredis"
slaveof 192.168.11.202 6379

  三、節點啓動

    以下相關路徑須要根據本身的狀況進行改動,可能和個人不同

[root@slave1 master_slave]# cd /opt/redis-3.0.0/master_slave/
[root@slave1 master_slave]# ./../src/redis-server redis-6379.conf 
[root@slave1 master_slave]# ./../src/redis-server redis-6380.conf 
[root@slave1 master_slave]# ./../src/redis-server redis-6381.conf

    確認主從關係

[root@slave1 master_slave]# ./../src/redis-cli -h 192.168.11.202 -p 6379 -a myredis info replication
# Replication
role:master
connected_slaves:2
slave0:ip=192.168.11.202,port=6380,state=online,offset=393,lag=0
slave1:ip=192.168.11.202,port=6381,state=online,offset=393,lag=0
master_repl_offset:393
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:2
repl_backlog_histlen:392

    還能夠從從節點視角來看:[root@slave1 master_slave]# ./../src/redis-cli -h 192.168.11.202 -p 6380 -a myredis info replication

  如若進行順利,按如上配置,一主二從環境搭建完畢

  四、從節點做用

    主要兩個做用:一、做爲主節點的一個備份,一旦主節點出現故障,從節點能夠做爲後備"頂"上來,而且保證數據儘可能不丟失(主從複製是最終一致性);二、從節點能夠拓展主節點的能力,一旦主節點不能支撐大併發的讀操做,從節點能夠在必定程度上幫助主節點分擔讀壓力

  五、主從複製問題

    一、一旦主節點出現故障,須要手動將一個從節點晉升爲主節點,同時須要修改應用方的主節點地址,還須要命令其餘從節點去複製新的主節點,整個過程須要人工干預

    二、主節點的寫能力受到單機的限制

    三、主節點的存儲能力受到單機的限制

redis sentinel部署

  此片主要講經過sentinel解決上述問題1,問題二、3則在下篇博客進行說明

  一、3個sentinel節點的配置文件基本一致,區別的只是端口

    sentinel-26379.conf

port 26379
daemonize yes
logfile "26379.log"
dir "/opt/soft/redis/data"
sentinel monitor mymaster 192.168.11.202 6380 2
#redis數據master節點設置了認證,則須要以下配置
sentinel auth-pass mymaster myredis
sentinel down-after-milliseconds mymaster 30000
sentinel parallel-syncs mymaster 1
sentinel failover-timeout mymaster 180000

    sentinel-26380.conf

port 26380
daemonize yes
logfile "26379.log"
dir "/opt/soft/redis/data"
sentinel monitor mymaster 192.168.11.202 6380 2
#redis數據master節點設置了認證,則須要以下配置
sentinel auth-pass mymaster myredis
sentinel down-after-milliseconds mymaster 30000
sentinel parallel-syncs mymaster 1
sentinel failover-timeout mymaster 180000

    sentinel-26381.conf

port 26381
daemonize yes
logfile "26379.log"
dir "/opt/soft/redis/data"
sentinel monitor mymaster 192.168.11.202 6380 2
#redis數據master節點設置了認證,則須要以下配置
sentinel auth-pass mymaster myredis
sentinel down-after-milliseconds mymaster 30000
sentinel parallel-syncs mymaster 1
sentinel failover-timeout mymaster 180000

  二、啓動sentinel

[root@slave1 master_slave]# ./../src/redis-sentinel sentinel-26379.conf 
[root@slave1 master_slave]# ./../src/redis-sentinel sentinel-26380.conf 
[root@slave1 master_slave]# ./../src/redis-sentinel sentinel-26381.conf

  三、sentinel確認

[root@slave1 master_slave]# ./../src/redis-cli -h 192.168.11.202 -p 26379 info Sentinel
# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
master0:name=mymaster,status=ok,address=192.168.11.202:6379,slaves=2,sentinels=3

  Redis  Sentinel 最終拓撲結構

 

  四、高可用測試

    手動停掉6379實例或者kill掉6379實例

[root@slave1 master_slave]# ./../src/redis-cli -h 192.168.11.202 -p 6379 -a myredis
192.168.11.202:6379> shutdown

    查看26379.log

    master節點已經自動切換到192.168.11.202:6380了,重啓6379,6379則是6378的從節點了。

[root@slave1 master_slave]# ./../src/redis-cli -h 192.168.11.202 -p 6380 -a myredis
192.168.11.202:6380> info replication
# Replication
role:master
connected_slaves:2
slave0:ip=192.168.11.202,port=6381,state=online,offset=80802,lag=1
slave1:ip=192.168.11.202,port=6379,state=online,offset=80802,lag=1
master_repl_offset:80945
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:2
repl_backlog_histlen:80944 

客戶端(Jedis)鏈接

  基於spring-boot開發,spring-boot-test測試, 這二者本文不作說明,網上資料不少,不熟悉的自行去補充; 工程結構以下圖

  pom.xml

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>

    <groupId>com.lee</groupId>
    <artifactId>redis</artifactId>
    <version>0.0.1-SNAPSHOT</version>
    <packaging>jar</packaging>

    <name>redis</name>
    <url>http://maven.apache.org</url>

    <properties>
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    </properties>

    <parent>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-parent</artifactId>
        <version>1.5.9.RELEASE</version>
    </parent>

    <dependencies>
        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-data-redis</artifactId>
        </dependency>
        
        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-test</artifactId>
            <scope>test</scope>
        </dependency>
        
        <!-- 日誌過濾器 -->
        <dependency>
            <groupId>org.codehaus.janino</groupId>
            <artifactId>janino</artifactId>
        </dependency>
    </dependencies>
</project>
View Code

  redis-sentinel.properties

redis.masterName=mymaster
redis.sentinels=192.168.11.202:26379,192.168.11.202:26380,192.168.11.202:26381
redis.timeout=10000
#鏈接master須要用到的密碼,若是redis數據節點開啓了鏈接認證
redis.password=myredis

# 鏈接池
# 鏈接池最大鏈接數(使用負值表示沒有限制)
redis.pool.maxActive=150
# 鏈接池中的最大空閒鏈接
redis.pool.maxIdle=10
# 鏈接池中的最小空閒鏈接
redis.pool.minIdle=1
# 獲取鏈接時的最大等待毫秒數,小於零:阻塞不肯定的時間,默認-1
redis.pool.maxWaitMillis=3000
# 每次釋放鏈接的最大數目
redis.pool.numTestsPerEvictionRun=50
# 釋放鏈接的掃描間隔(毫秒)
redis.pool.timeBetweenEvictionRunsMillis=3000
# 鏈接最小空閒時間(毫秒)
redis.pool.minEvictableIdleTimeMillis=1800000
# 鏈接空閒多久後釋放, 當空閒時間>該值 且 空閒鏈接>最大空閒鏈接數 時直接釋放(毫秒)
redis.pool.softMinEvictableIdleTimeMillis=10000
# 在獲取鏈接的時候檢查有效性, 默認false
redis.pool.testOnBorrow=true
# 在空閒時檢查有效性, 默認false
redis.pool.testWhileIdle=true
# 在歸還給pool時,是否提早進行validate操做
redis.pool.testOnReturn=true
# 鏈接耗盡時是否阻塞, false報異常,ture阻塞直到超時, 默認true
redis.pool.blockWhenExhausted=true
View Code

  RedisConfig.java

package com.lee.redis.config;

import java.util.Arrays;
import java.util.HashSet;
import java.util.Set;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.context.annotation.PropertySource;

import redis.clients.jedis.JedisPoolConfig;
import redis.clients.jedis.JedisSentinelPool;

@Configuration
@PropertySource("redis/redis-sentinel.properties")
public class RedisConfig {
    private static final Logger LOGGER = LoggerFactory.getLogger(RedisConfig.class);
    
    @Value("${redis.masterName}")
    private String masterName;
    @Value("${redis.sentinels}")
    private String sentinels;
    @Value("${redis.timeout}")
    private int timeout;
    @Value("${redis.password}")
    private String password;
    
    @Value("${redis.pool.maxActive}")
    private int maxTotal;

    @Value("${redis.pool.maxIdle}")
    private int maxIdle;

    @Value("${redis.pool.minIdle}")
    private int minIdle;

    @Value("${redis.pool.maxWaitMillis}")
    private long maxWaitMillis;

    @Value("${redis.pool.numTestsPerEvictionRun}")
    private int numTestsPerEvictionRun;

    @Value("${redis.pool.timeBetweenEvictionRunsMillis}")
    private long timeBetweenEvictionRunsMillis;

    @Value("${redis.pool.minEvictableIdleTimeMillis}")
    private long minEvictableIdleTimeMillis;

    @Value("${redis.pool.softMinEvictableIdleTimeMillis}")
    private long softMinEvictableIdleTimeMillis;

    @Value("${redis.pool.testOnBorrow}")
    private boolean testOnBorrow;

    @Value("${redis.pool.testWhileIdle}")
    private boolean testWhileIdle;

    @Value("${redis.pool.testOnReturn}")
    private boolean testOnReturn;

    @Value("${redis.pool.blockWhenExhausted}")
    private boolean blockWhenExhausted;
    
    @SuppressWarnings({ "unchecked", "rawtypes" })
    @Bean
    public JedisSentinelPool jedisSentinelPool(JedisPoolConfig poolConfig) {
        LOGGER.info("sentinel set : {}", sentinels);
        Set sentinelSet = new HashSet(Arrays.asList(sentinels.split(",")));
        JedisSentinelPool jedisPool = new JedisSentinelPool(masterName, sentinelSet, poolConfig, timeout, password);
        return jedisPool;
    }
    
    @Bean
    public JedisPoolConfig jedisPoolConfig() {
        JedisPoolConfig jedisPoolConfig = new JedisPoolConfig();
        jedisPoolConfig.setMaxTotal(maxTotal);
        jedisPoolConfig.setMaxIdle(maxIdle);
        jedisPoolConfig.setMinIdle(minIdle);
        jedisPoolConfig.setMaxWaitMillis(maxWaitMillis);
        jedisPoolConfig.setNumTestsPerEvictionRun(numTestsPerEvictionRun);
        jedisPoolConfig
                .setTimeBetweenEvictionRunsMillis(timeBetweenEvictionRunsMillis);
        jedisPoolConfig
                .setMinEvictableIdleTimeMillis(minEvictableIdleTimeMillis);
        jedisPoolConfig
                .setSoftMinEvictableIdleTimeMillis(softMinEvictableIdleTimeMillis);
        jedisPoolConfig.setTestOnBorrow(testOnBorrow);
        jedisPoolConfig.setTestWhileIdle(testWhileIdle);
        jedisPoolConfig.setTestOnReturn(testOnReturn);
        jedisPoolConfig.setBlockWhenExhausted(blockWhenExhausted);

        return jedisPoolConfig;
    }
}
View Code

  Application.java

package com.lee.redis;

import org.springframework.boot.Banner;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.EnableAutoConfiguration;
import org.springframework.context.annotation.ComponentScan;
import org.springframework.context.annotation.Configuration;

@Configuration
@EnableAutoConfiguration
@ComponentScan
public class Application {

    public static void main(String[] args) {
        
        SpringApplication app = new SpringApplication(Application.class);
        app.setBannerMode(Banner.Mode.OFF);            // 是否打印banner
        // app.setApplicationContextClass();        // 指定spring應用上下文啓動類
        app.setWebEnvironment(false);
        app.run(args);
    }
}
View Code

  RedisTest.java

package com.lee.redis;

import org.junit.Test;
import org.junit.runner.RunWith;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.test.context.SpringBootTest;
import org.springframework.test.context.junit4.SpringRunner;

import redis.clients.jedis.Jedis;
import redis.clients.jedis.JedisSentinelPool;

@RunWith(SpringRunner.class)
@SpringBootTest(classes = Application.class)
public class RedisTest {

    private static final Logger LOGGER = LoggerFactory.getLogger(RedisTest.class);
    
    @Autowired
    private JedisSentinelPool sentinelPool;
    
    @Test
    public void getNameTest() {
        Jedis jedis = null;
        try {
            jedis = sentinelPool.getResource();
            String name = jedis.get("name");
            System.out.println("name is " + name);
        } catch(Exception e) {
            LOGGER.error(e.getMessage(), e);
        } finally {
            if (jedis != null) {
                jedis.close();
            }
        }
        
    }
    
}
View Code

  更多詳情請上個人github

  運行RedisTest.java的getNameTest方法(name屬性已經在redis中設置,沒設置的須要提早設置),獲得結果:

注意點

  一、有人可能會有這樣的疑問:爲何經過sentinel來獲取redis的鏈接,而不是直接鏈接master來獲取redis鏈接呢?

    試想一下,客戶端直接經過master節點獲取redis鏈接,若是master節點掛掉了,雖然Redis Sentinel能夠完成故障轉移,可是客戶端沒法獲取這個變化,那麼客戶端就沒法獲取redis鏈接了;

    最瞭解master節點信息的就是Sentinel節點集合,因此經過sentinel來獲取redis鏈接就能知足高可用的要求了。

  二、redis master的故障轉移不影響客戶端鏈接代碼, 可是轉移期間內,經過sentinel是獲取不到主節點的鏈接的, 由於轉移期間內master節點還沒被選舉出來;

參考

  《Redis開發與運維》

  http://www.redis.cn/topics/sentinel.html

相關文章
相關標籤/搜索