Redis6 cluster setup

Redis6 Cluster building

At present redis The latest version is already 6.2.6, On cluster construction and redis3.x、redis4.x Difference is very big .redis5 in the future , There's no need to install ruby 了 , And the new version redis Creating a cluster is no longer used redis-trib.rb, While using redis-cli. The following is the new version redis6.2.6 Cluster setup record . This article uses three servers ,6 individual redis Service instance （redis The cluster will automatically try to prevent two instances of the same machine from being master-slave to each other ）.

---

One 、 Environmental preparation

ip	port
101.0.0.101	7001 、7002
101.0.0.102	7001 、7002
101.0.0.103	7001 、7002

Two 、 Build and start redis example

1.1 take redis-6.2.6.tar.gz( Download address ：https://redis.io/download) Upload to server , Decompress and compile

yum -y install gcc

tar xzf redis-6.2.6.tar.gz -C home/

cd home/redis-6.2.6

make MALLOC=libc 

make PREFIX=/home/redis-6.2.6  install  # Such as make PREFIX=/usr/local/redis install, keyword  PREFIX=  This keyword is used to specify the path where the program is stored at compile time , Suppose you don't add the keyword Linux The executable file will be stored in /usr/local/bin Catalog , Library files will be stored in /usr/local/lib Catalog . The configuration file will be stored in /usr/local/etc Catalog .
 Other resource files will be stored in usr/local/share Catalog . The directory number is specified here to facilitate subsequent unloading , Follow up direct rm -rf /usr/local/redis  You can delete redis.

1.2 Create related directories , Modify the configuration file

mkdir home/rediscluster/{ 
    7000,7001} -p

mkdir home/rediscluster/7000/{ 
    conf,data,log} -p

mkdir /home/rediscluster/7001/{ 
    conf,data,log} -p

cp /home/redis-6.2.6/redis.conf  /home/rediscluster/7000/conf/

cp /home/redis-6.2.6/redis.conf  /home/rediscluster/7001/conf/

1.3 Modify the configuration file

modify /home/rediscluster/7000/conf/redis.conf（ The following cluster configuration must be modified , Other configurations are modified according to the actual situation ）：

bind 101.0.0.101

port 7000

daemonize yes

pidfile "/home/rediscluster/7000/data/redis_7000.pid"

logfile "/home/rediscluster/7000/log/redis_7000.log"

dbfilename dump-7000.rdb

dir "/home/rediscluster/7000/data/"

masterauth "myredis"   # Set the password ,redis Be sure to enable password authentication requirepass and masterauth Simultaneous setting ;masterauth effect ： Mainly aimed at master Corresponding slave Node settings , stay slave Node data synchronization is used .requirepass effect ： Restrict login permissions ,redis For each node requirepass Can be independent 、 Different .requirepass Verify client ,masterauth Verify slave Library .
requirepass "myredis"  

cluster-enabled yes

cluster-node-timeout 5000

take 7000 Copy the configuration file of the directory to 7001 Instance and pay attention to the 7000 Change it to 7001：

cp /home/rediscluster/7000/conf/redis.conf  /home/rediscluster/7001/conf/redis.conf

sed -i 's#7000#7001#g' /home/rediscluster/7001/conf/redis.conf

1.4 Configure environment variables

echo 'export REDIS_HOME=/home/redis-6.2.2' >> /etc/profile 

echo 'export PATH=$REDIS_HOME/bin:$PATH' >> /etc/profile

source /etc/profile

1.5 start-up redis

redis-server /home/rediscluster/7000/conf/redis.conf

redis-server /home/rediscluster/7001/conf/redis.conf

1.6 The other two servers

Repeat the above steps on the other two servers , Pay attention to redis.conf Inside bind IP Change to native IP.
Ensure the availability of three servers redis Node starts normally ：

3、 ... and 、 establish redis colony

On three servers redis After the instance is successfully started , Execute the command on the first server to create the cluster （ Note that if redis.conf The password is set in , Here we need to add -a ‘ password ’, If there is no password , No more -a Parameters ）：

 Password ：
redis-cli --cluster create --cluster-replicas 1 101.0.0.101:7000  101.0.0.101:7001  101.0.0.102:7000  101.0.0.102:7001  101.0.0.103:7000  101.0.0.103:7001 -a myredis
 password-less ：
redis-cli --cluster create --cluster-replicas 1 101.0.0.101:7000  101.0.0.101:7001  101.0.0.102:7000  101.0.0.102:7001  101.0.0.103:7000  101.0.0.103:7001

/usr/local/redis/bin/redis-cli --cluster create 10.11.0.148:7001 10.11.0.149:7002 10.11.0.149:7001 10.11.0.150:7002 10.11.0.150:7001 10.11.0.148:7002 --cluster-replicas 1

Create a screenshot of the process , Note the confirmation input in the figure “yes”.

Four 、 verification redis colony

After the cluster is created successfully , Sign in redis verification , Pay attention to adding cluster parameters -c：

redis-cli -h 101.0.0.101 -p 7000 -a myredis -c  # Sign in redis colony 

101.0.0.101:7000> cluster nodes  # View the cluster nodes 

101.0.0.101:7000> cluster info   # View cluster information 

verification OK,redis6.2.2 The cluster has been set up .

5、 ... and 、Redis Configuration file parameter description

1. Redis The default is not to run as a daemon , This configuration item can be modified , Use yes Enable daemons 

daemonize no

2.  When Redis Run as a daemon ,Redis By default the pid write in /var/run/redis.pid file , Can pass pidfile Appoint 

pidfile /var/run/redis.pid

3. Appoint Redis Listening port , The default port is 6379, In one of his blog posts, the author explains why 6379 As default port , because 6379 On the phone buttons MERZ The corresponding number , and MERZ From the Italian singer Alessia Merz Name 

port 6379

4.  The bound host address 

bind 127.0.0.1

5. When   How long does the client sit idle before closing the connection , If specified as 0, Indicates that the feature is turned off   

timeout 300
   
6.  Specifies the logging level ,Redis There are four levels supported ：debug、verbose、notice、warning, The default is verbose

loglevel verbose

7.  Logging mode , The default is standard output , If configured Redis Run for daemon mode , This is configured for logging as standard output , The log will be sent to /dev/null

logfile stdout

8.  Set the number of databases , The default database is 0, have access to SELECT <dbid> The command specifies the database on the connection id

databases 16

9.  Specify how long , How many update operations are there , Synchronize the data to the data file , Multiple conditions can be used 

save <seconds> <changes>

Redis Three conditions are provided in the default configuration file ：

save 900 1

save 300 10

save 60 10000

 respectively 900 second （15 minute ） There are 1 A change ,300 second （5 minute ） There are 10 Changes and 60 In seconds 10000 A change .

10.  Specifies whether data is compressed when stored to a local database , The default is yes,Redis use LZF Compress , If in order to save money CPU Time , This option can be turned off , But it can make the database files huge 

rdbcompression yes

11.  Specifies the local database file name , The default value is dump.rdb

dbfilename dump.rdb

12.  Specifies the local database hosting directory 

dir ./

13.  Set when the native is slav The service , Set up master Service IP Address and port , stay Redis Startup time , It will automatically follow master Data synchronization 

slaveof <masterip> <masterport>

14.  When master When the service is password protected ,slav Service connection master Password 

masterauth <master-password>

15.  Set up Redis Connect the password , If the connection password is configured , The client is connecting Redis Need to go through AUTH <password> Command provide password , Off by default 

requirepass foobared

16.  Sets the maximum number of client connections at one time , Default limit ,Redis The number of client connections that can be opened simultaneously is Redis The maximum number of file descriptors that a process can open , If you set  maxclients 0, No restrictions . When the number of client connections reaches the limit ,Redis The new connection is closed and returned to the client max number of clients reached error message 

maxclients 128

17.  Appoint Redis Maximum memory limit ,Redis Data is loaded into memory at startup , After reaching the maximum memory ,Redis It will first try to clear those that are due or are about to expire Key, When this method is handled   after , Still reaches the maximum memory setting , Will no longer be able to write , But you can still do a read .Redis new vm Mechanism , Will be able to Key Storage memory ,Value Will be stored in a swap District 

maxmemory <bytes>

18.  Specifies whether to log after each update operation ,Redis By default, data is written asynchronously to disk , If you don't turn it on , Data may be lost for a period of time during a power outage . because  redis The itself synchronization data file is pressed above save Condition to synchronize , So some data will only exist in memory for a while . The default is no

appendonly no

19.  Specifies the file name of the update log , The default is appendonly.aof

 appendfilename appendonly.aof

20.  Specifies an update log condition , share 3 Optional values ： no： Means that the operating system synchronizes the data cache to disk （ fast ） always： Represents a manual call after each update operation fsync() Write data to disk （ slow , Security ） everysec： That means once per second （ compromise , The default value is ）

appendfsync everysec

21. Specifies whether the virtual memory mechanism is enabled , The default value is no, A brief introduction ,VM The mechanism stores data in pages , from Redis Cold data for less visited pages swap To disk , Pages with multiple accesses are automatically swapped out from disk to memory （ I will analyze it carefully in the following articles Redis Of VM Mechanism ）

vm-enabled no

22.  Virtual memory file path , The default value is /tmp/redis.swap, Not more than one Redis The instance Shared 

 vm-swap-file /tmp/redis.swap

23.  I'm going to make everything greater than vm-max-memory Is stored in virtual memory , No matter what vm-max-memory Set up small , All index data is stored in memory (Redis Index data of   Namely keys), in other words , When vm-max-memory Set to 0 When , It's all value Both exist on disk . The default value is 0

 vm-max-memory 0

24. Redis swap The file is divided into many parts page, An object can be stored in more than one page above , But a page Can't be Shared by multiple objects ,vm-page-size It's based on storage   The data size is set , The authors suggest storing many small objects ,page The size is best set to 32 perhaps 64bytes; If you store large objects , You can use a larger one page, If you don't   determine , Just use the default values 

 vm-page-size 32

25.  Set up swap In the document page Number , Because the page table （ One that indicates that a page is free or in use bitmap） It's in memory ,, Each on disk 8 individual pages Will consume 1byte Of memory .

 vm-pages 134217728

26. Set access swap Number of threads in the file , It is best not to exceed the audit of the machine , If set to 0, So all right swap The operation of the file is serial , May cause a relatively long delay . The default value is 4

 vm-max-threads 4

27.  Set when to reply to the client , Whether to combine smaller packages into a single package , On by default 

glueoutputbuf yes

28.  Specifies when a certain number or maximum of elements exceeds a certain threshold , A special hash algorithm is used 

hash-max-zipmap-entries 64

hash-max-zipmap-value 512

29.  Specifies whether to activate the reset hash , On by default （ In the introduction Redis The hash algorithm is introduced in detail ）

activerehashing yes

30.  Specifies that additional configuration files are included , More than one can be on the same host Redis The same configuration file is used between instances , At the same time, each instance has its own specific configuration file 

include /path/to/local.conf