Summary of redis basic knowledge points

Redis Default port 6379

One 、 install

1. Go to redis.cn Check out the latest stable download address ,http://download.redis.io/releases/redis-6.0.6.tar.gz

2. download wget http://download.redis.io/releases/redis-6.0.6.tar.gz

3. decompression tar -zxvf redis-6.0.6.tar.gz

4. Move Put it in /usr/local/redis Under the table of contents sudo mv redis-6.0.6 /usr/local/redis/ because redis directory does not exist , So it is equivalent to renaming

5. Get into redis Catalog cd /usr/local/redis-6.0.6

6. Generate sudo make because redis yes C language-written , This step is equivalent to compiling

​ If something goes wrong , Probably not installed gcc and make, After installation make distclean In addition to compiling the content, recompile *

7. test sudo make test

8. install sudo make install take redis The command to install to /usr/local/bin/ Catalog

9. Get into /usr/local/bin View in cd /usr/local/bin ls -all

10. Copy the configuration file to /etc/ Under the table of contents The configuration file is /usr/local/redis/redis.conf

​ sudo cp /usr/local/redis/redis.conf /etc/redis/

Two 、 Basic use

Start server

redis-server Start server

ps aux|grep redis See if it starts

redis-server --help View help for startup

sudo redis-server /etc/redis/redis.conf Specify the loaded profile

Start client

redis-cli Start client

redis-cli --help View startup help , Parameters can be added （ host 、 Ports, etc ）

redis-cli -h ip Address -p Port number

ping

After starting the client, enter ping If the result is PONG Indicates that the client has successfully connected redis Server side

select

Switch database

select Numbers The range of numbers is 0-databases Maximum minus 1

The default is in select 0

help

help Command name View help for using this command

Core configuration options

The configuration information is in /etc/redis/redis.conf

bind 127.0.0.1 If you need remote access , This line can be annotated or bound to a real ip

port 6379 Port number , Default 6379

daemonize yes Set to run as a daemon ,no Is not a daemon

​ If running as a daemons , Will not block on the command line , Similar to service , To ensure the redis The stable operation of the service in the background

​ If running as a non daemon , The current terminal is blocked

databases 16 redis The default is 16 A database

logfile Log path

dbfilename dump.rdb Data files , It is stored in the home directory by default

dir /var/lib/redis Data file storage path

slaveof Master slave copy , Similar to dual backup

3、 ... and 、 Basic data structure

redis yes key:value Data structure of , Each data is a key value pair

Character type of key type , And can't repeat

Value types can have certain data structures , It can be a string string、 list list、 Hash hash、 aggregate set、 Ordered set zset

Four 、 Additions and deletions

string type

stay redis Of string type value The maximum length of data that can be held is 512M

set Set up

set name LiKang

set age 22

get obtain

get name

get age

del Delete

del name

del age

Set when inserting key The expiration time of

setex key Number of seconds value

ttl key Check out this key The validity of the ,-1 For permanent existence

mset Set up multiple key-value

mset name Liang age 22 Be careful ： If... Is set later name Before name The value of will be overridden

mget key1 key2 Get multiple values

append Additional

append key value

Key command

del key1 key2 Delete this key The key/value pair

keys * See what they have key

keys a* see a At the beginning key

exists key see key Whether there is ,1 There is ,0 There is no

type key Check out this key Of value The type of

expire key Number of seconds to key Set expiration time

hash type

key The value of is attribute : value , Such key value pairs , The property value is string type

hset key field value Set up

​ for example hset person name LiKang

hget key field obtain value value

​ for example hget person name

hmset key field1 value1 field2 value2 Set multiple values

hmget key field1 field2 Get multiple values

hgetall key Access to this key All key value pairs under

hkeys key Access to this key All properties under

hvals key Access to this key All attribute values under

hdel key field1 field2 Delete a property

del key Delete this key The key/value pair

If it is updated, please hset key field value Just insert , The same field will automatically overwrite the original data to realize the update

list type

The list element type is string, And the data is sorted according to the insertion order , Elements can be repeated

lpush Insert... From the left

​ lpush key value1 value2 The return value is key in value The number of

​ lrange key num1 num2 Get from left to right num1 To num2 The element values between （ Include num2）,0 Represents the first element ,-1 Represents the last element ,l representative list, No left

rpush Insert... From the right

​ rpush key value1 value2 The return value is key in value The number of

​ nothing rrange command

lrem key count value Remove data ,l representative list, No left

​ count：0 On behalf of the delete key All values in are value The elements of

​ count>0 Delete the left value as value The elements of , The number of deletions is count

​ count<0 Delete the value on the right as value The elements of , The number of deletions is count

lset key Subscript value Update data

linsert key before(after) value1 value2 stay value1 In front of （ Back ） Insert value2

set type

unordered set , Element is string type , Element is unique , No repetition , There is no modification operation for the collection

sadd key member1 member2 Add data

smembers key get data

srem key member Remove data

zset type

Ordered set , Element type is string, Element is unique , No repetition , Each element is associated with a double Type of score, Said the weight , Sort elements from small to large by weight , And there is no set modification operation

zadd key score1 member1 score2 member2 Add data

​ If score The value of already exists , The new element will be inserted into the score After or before elements with the same value

​ If member The value of already exists , It will be modified socre The value of is the new value

zrange key num1 num2 obtain num1 To num2 Data between （ Include num2）

zrangeby score key min max obtain score Values in min To max Member between （ Include min and max）

zscore key member return member Of score value

zrem key member1 member2 Deletes the specified element

zremrangebyscore key min max Delete score Values in min To max Elements in scope （ Include min and max）

5、 ... and 、 Build master slave

One master You can have multiple slave, One slave You can have more than one slave, Go on like this , Form a powerful multi-level server cluster architecture .

master To write data ,slave To read data , Statistics , The read-write ratio of the website is 10:1, That is to say 11 Only one of the users is writing data .

Read and write separation can be realized through master-slave configuration .

Configure the primary server

View the current host's ip Address ifconfig

modify /etc/redis/redis.conf file bind Current host ip

restart redis service

​ sudo service redis stop

​ sudo redis-server redis.conf

Configure slave

Copy /etc/redis/redis.conf file sudo cp redis.conf ./slave.conf

modify slave.conf file

​ bind host ip （ If it is a different computer from the main server , Bind native ip that will do ）

​ port 6378

​ Add a row slaveof Current host ip Primary server port （ Default 6379）

start-up redis service sudo redis-server slave.conf

View master-slave relationship redis-cli -h Current host ip info Replication

PS： The master server can read and write , You can only read from the server , And the master-slave server data can be displayed synchronously

start-up redis-cli -h host ip -p Port number

6、 ... and 、 Set up the cluster

Although one master can follow more , But if the number of simultaneous visits is too large （1000w）, The main service will be dead for sure ; Access to natural disasters , The data service is down

Big companies have a lot of servers .

The concept of cluster

A cluster is a group of independent , A computer connected by a high-speed network , They form a group , And managed in a single system model . When a client interacts with a cluster , Clustering is like a stand-alone server , Cluster configuration is used to improve availability and scalability . When the request arrives, it is first handled by the load balancing server , Forward the request to another server .

classification

Software level ： There's only one computer , On this computer, several redis service

Hardware level ： Computers with multiple entities , There's one on every computer redis Or more redis service

Cluster building

At one ubuntu Build a cluster on the machine

Machine configuration 1

stay Desktop Create a folder in the directory “redis_cluster", Create... Under this folder 7000.conf, The contents are as follows

#  Port number 
port 7000
#  host ip
bind 192.168.37.129
#  Whether to guard （ backstage ） Run in process mode 
daemonize yes
# pid file 
pidfile 7000.pid
#  Whether to use cluster 
cluster-enabled yes
#  Cluster files 
cluster-config-file 7000_node.conf
#  The timeout of the cluster 
cluster-node-timeout 15000
#  Backup related 
appendonly yes

stay ”redis_cluster“ Create under folder 7001.conf、7002.conf, except 7000 Change it , The content is the same as above

Machine configuration 2

And configure the machine 1 equally , stay Desktop Create under directory redis_cluster Folder , It creates 7003.conf、7004.conf、7005.conf, In addition to putting the host ip and 7000 Change it , Everything else is the same .

But there is only one machine here , So it will also 7003.conf、7004.conf、7005.conf Two files on the machine 1 On

Create clusters

First in redis_cluster Start under folder redis service

sudo redis-server 7000.conf
sudo redis-server 7001.conf
sudo redis-server 7002.conf
sudo redis-server 7003.conf
sudo redis-server 7004.conf
sudo redis-server 7005.conf

redis The installation package for contains redis-trib.rb, For creating clusters

The next operation is on the machine 1 on

Copy the command , This allows you to call this command in any directory

sudo cp /usr/share/doc/redis-tools/examples/redis-trib.rb /usr/local/bin/

redis 5.0 Previous versions created clusters

install ruby Environmental Science , because redis-trib.rb Yes, it is ruby Developed

sudo apt-get install ruby

Create clusters （ Must be on first redis service ）

–replicas 1 Indicates that the cluster mode created is “ A master from ”, There are three nodes （Node）

redis-trib.rb create --replicas 1 192.168.37.129:7000 192.168.37.129:7001 192.168.37.129:7002 192.168.37.129:7003 192.168.37.129:7004 192.168.37.129:7005

redis 5.0 Later versions create clusters

5.0 Version discards redis-trib

Run the following command directly ,–cluster-replicas 1 Indicates that the cluster mode created is “ A master from ”, There are three nodes （Node）

redis-cli --cluster create 192.168.37.129:7000 192.168.37.129:7001 192.168.37.129:7002 192.168.37.129:7003 192.168.37.129:7004 192.168.37.129:7005 --cluster-replicas 1

（0 slots） by slave service , No 0 For the master service （redis 5.0 In previous versions, no matter slave still master Are shown as master）

Connect to cluster

redis-cli -h host ip -p Port number -c

get key get data , If the data is not in the current node, it will be redirected automatically

Cluster considerations

When inserting data , If the connection is from the server , Will pass CRC16 The algorithm redirects to another master server for insertion .

If the primary server is connected , Also through CRC16 The algorithm redirects to another master server for insertion .

Every node in the cluster has an equal relationship , Each node saves its own data and the status of the entire cluster , Each node is connected to all other nodes , And those connections stay active , This ensures that we only need to connect to any node in the cluster , You can get data from other nodes .

The cluster will store the data in a master node , And then here master Corresponding to it slave Data synchronization between , Only when one master After hanging up , To start a corresponding slave node , act as master

There must be three or more master nodes , Otherwise, it will fail to create the cluster , And when the number of surviving primary nodes is less than half of the total nodes , The entire cluster will not be able to provide services .