Shulou(Shulou.com)06/01 Report--

1. Redis master-slave building

1. Download and decompress

Yum install-y gcc gcc-c++ pcre zlib pcre-devel tcl wget http://download.redis.io/releases/redis-3.2.4.tar.gztar-zxvf redis-3.2.4.tar.gzcd redis-3.2.4make cd src & & make test & & make installmkdir / etc/rediscp.. / redis.conf / etc/redis/redis.conf

two。 System parameter tuning

Vim / etc/sysctl.confnet.core.somaxconn = 20480 # maximum queue length to handle sudden large concurrent connection requests. Default is 128net.ipv4.tcp_max_syn_backlog = 20480 # semi-connection queue length. This value is limited by memory size. The default value of 1024vm.overcommit_memory = 10 means to check whether enough memory is available. If so, allocation is allowed. If there is not enough memory, the request is rejected and an error is returned to the application. 1 allows requests to allocate more than physical memory plus swap memory 2 the kernel always returns truesysctl-p # load effective parameters

3. Profile description

3.1 running in the background

Daemonize yes

3.2 bind address snooping

The default bind is 127.0.0.1. This configuration allows only local access to bind 0.0.0.0.

3.3 Log configuration

Logfile "/ var/log/redis.log"

3.4 Service startup script

Vim / etc/init.d/redis# chkconfig: 2345 90 1 in start description: service of redis for start and stop add by tomener PATH=/usr/local/bin:/sbin:/usr/bin:/binREDISPORT=6379EXEC=/usr/local/bin/redis-serverREDIS_CLI=/usr/local/bin/redis-cli PIDFILE=/var/run/redis_6379.pidCONF= "/ etc/redis/redis.conf" AUTH= "Passwd" BIND_IP='0.0.0.0'case "$1" in start) If [- f $PIDFILE] then echo "$PIDFILE exists Process is already running or crashed. " Else echo "Starting Redis server..." $EXEC $CONF fi if ["$?" = "0"] then echo "Redis is running..." Fi; stop) if [!-f $PIDFILE] then echo "$PIDFILE exists, process is not running." Else PID=$ (cat $PIDFILE) echo "Stopping..." $REDIS_CLI-h $BIND_IP-a $AUTH-p $REDISPORT SHUTDOWN sleep 2 while [- x $PIDFILE] do echo "Waiting for Redis to shutdown..." Sleep 1 done echo "Redis stopped" fi;; restart | force-reload) ${0} stop ${0} start *) echo "Usage: / etc/init.d/redis {start | stop | restart | force-reload}" > & 2 exit 1esacchkconfig redis on # set boot self-startup service redis restart # startup service

3.5 set password

Edit the configuration file / etc/redis/redis.conf. Restart takes effect.

Requirepass redispwd redis-cli-p 6379-a redispwd # requires-a parameter to enter a password to connect to the database

3.6 backup recovery

Use the command to view the backup directory, which is the directory at the start of the service. Restore can store files in this directory, restart and start the redis service to restore the number of rdb.

CONFIG GET dir # View backup directory save # perform backup

4.Redis persistence

4.1 rdb mode

How it works:

When redis generates a dump.rdb file, the working process is as follows:

The redis main process fork a child process that comes out of fork to dump the data set of memory into a temporary RDB

When the child process has finished writing to the temporary RDB file, redis replaces the old RDB file with the new RDB file

The default configuration is as follows:

Save 900 1save 300 10save 60 10000

Its significance:

Save the data to the hard disk every 900s when 1 key updates the value

Save to hard disk every 300 seconds when 10 key update values

Save to hard disk every 60 seconds when 10000 key update values

4.2 aof mode

Aof is essentially a redis operation (write operation) log file. Aof is not enabled by default and needs to be set in the configuration file. Just change this line to appendonly yes in the configuration file.

How it works:

AOF: append only file .

Whenever Redis executes a command that changes the dataset, the command is appended to the end of the AOF file.

When redis restarts, the program can recover the data through the AOF file.

There are three appedn methods:

Appendfsync alwaysappendfsync everysecappendfsync no

Appendfsync always executes fsync every time a new command is appended to the AOF file: very slow and very secure

Appendfsync everysec fsync once per second: fast enough (almost as fast as using RDB persistence), and only 1 second of data is lost in the event of a failure.

Appendfsync no never fsync: leave the data to the operating system for processing. Faster and less secure options.

The recommended (and default) measure is fsync per second, which is a fsync strategy that combines speed and security.

Aof can ensure the security of data, but it is time-consuming to restart, and the size of aof files is larger than rdb files.

5.Redis master-slave replication

5.1 Overview

Redis's replication mechanism allows slave to transfer copies from master to a full data backup over the network. It has the following characteristics:

Asynchronous replication, starting with version 2.8, slave can get data from master from time to time.

Allows multiple slave to be configured in a single master for easy scale-out

Slave allows other slave to connect to itself. A slave can connect to other slave in addition to master. To form a graphical structure.

Master is non-blocking during replication, which means that master can still process client requests during replication.

Slave is also non-blocking during replication and can accept requests from clients, but it uses previous old data. It can be configured to determine whether slave responds to client requests with old data during replication, and if configured to no, slave will return an error message to the client. However, when the new data is fully received, the new data must be replaced with the old data, that is, the old data must be deleted. In this time window of the replacement data, slave will reject the client's request and connection.

Replication is generally used to achieve scalability, for example, multiple slave can be configured as "read-only", or purely redundant backup of data.

You can use replication to prevent master from persisting the entire dataset to the hard disk each time it is persisted. Simply configure the master not to persist (comment out the configuration items related to persistence in the configuration file), connect to a slave, and the slave is configured with the persistence option. It is important to note, however, that in this scenario, you must ensure that master does not start automatically.

5.2 Security

Security of Master when Replication persistence is turned off when there is a need to use the replication mechanism, it is generally strongly recommended to turn on the master persistence switch. Even if you don't turn on the persistence switch in order to avoid the latency effects of persistence, you should configure master not to start automatically.

5.3 how replication works

If you configure a slave for master, regardless of whether the slave is connected to Master for the first time or not, it will send a SYNC command to master to request replication of data.

After receiving the SYNC command, master persists the data in the background, and during the persistence, master continues to receive requests from the client, caching these requests that may modify the dataset in memory. When the persistence is complete, master sends the dataset to slave,slave, persists the received data, and then loads it into memory. Master then sends the previously cached command to slave.

When the connection between master and slave is disconnected for some reason, slave can automatically reconnect the master. If master receives multiple slave concurrent connection requests, it will only persist once instead of one connection, and then send this persistent data to multiple concurrent connected slave.

When master and slave are disconnected and reconnected, the entire data is usually replicated. But starting with the redis2.8 version,

5.4 Master-Slave configuration

The configuration related to replication is relatively simple, just add the following line to the configuration file of slave:

Slaveof masterIPaddress 6379

If master sets the password through the requirepass configuration item, slave needs to verify the password for each synchronization operation, by adding the following configuration item to the slave configuration file:

Masterauth

6 commonly used commands

Start redis:

Redis-server / etc/redis/redis.conf

Client command tool

/ usr/local/bin/redis-cli-h 192.168.1.1192.168.1.1purl 6379 > auth Passwd # login Password Verification 192.168.1.1slave 6379 > info # View database status 192.168.1.1keys 6379 > info replication # View replication status of slave 192.168.1.1slave 6379 > set key 123 # insert data 192.168.1.1slave 6379 > keys * # list data flushdb # clear current data flushall # clear all databases

7. Test master-slave and switch

7.1 Master-Slave Test

7.2 Master-Slave switching

Stop the master

Switch from slave to master

Redis-cli-h localhost slaveof NO ONE

Restore the original master database

Overwrite the copy of the dump.rdb file under the current main redis root directory to the root directory of the original main redis

Start the original master redis

Switching from database to master database redis-cli-h localhost-p 6379 slaveof 172.17.10.193 6379

The test has been switched back.

2. Setting up Redis clusters

The higher version of Redis uses ruby to implement the cluster, so you need a ruby environment. After installing the ruby environment and redis's gem interface, you can use redis's redis-trib.rb script to create the cluster.

After the Redis installation is successful, create a separate directory

Mkdir / usr/local/redis/cluster-p

Then create directories 7000, 7001, 7002, 7003 under it

Mkdir / usr/local/redis/cluster/7000-pcp redis.conf / usr/local/redis/cluster/7000/

Copy redis.conf to these directories, and modify the port number port and directory name in redis.conf respectively.

Modify the configuration file separately

Daemonize yes / / redis background runs pidfile / var/run/redis_7000.pid / / pidfile file corresponding to 7000Magne7002Magne7003port 7000 / / Port 7000Power7002Power7003clusterlyenabled yes / / start the cluster and remove the cluster-config-file nodes_7000.conf / / cluster For the first time, the configuration file starts the automatic generation of 7000meme7001j7002clustermusic / color timeout 5000 / / the request timeout is set for 5 seconds enough to enable the appendonly yes / / aof log if necessary. It records a log for each write operation.

Check to see if the service has been up

two。 Install related software packages

Yum-y install ruby ruby-devel rubygems rpm-build

Then use the command gem to install the redis interface gem is a toolkit for ruby

Gem install redis

If the installation fails, install it manually

Wget https://rubygems.global.ssl.fastly.NET/gems/redis-3.2.1.gemsudo gem install-l. / redis-3.2.1.gem

3.redis-trib.rb script to create a cluster

Cd / root/redis-3.2.4/src./redis-trib.rb create-- replicas 1 172.17.10.191VG 7000 172.17.10.191VOL 7001 172.17.107005 172.17.10.191VOR 7006

-- replicas 1 means that each master node is automatically assigned a slave node.

. / redis-trib.rb check 172.17.10.191:7000

4. test

Redis-cli-c-p 7000

Set the data, and then stop the redis that port 7000 is listening to

Log in to others and check to get the data.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.