PV introduction

Million pv website architecture

Deployment environment

host name	ip address	role
master	192.168.0.106	Nginx + maintained main database, redis main database and mysql data main database
backup	192.168.0.115	Nginx + maintained backup, redis backup, mysql data slave
web1	192.168.0.104	Backend tomcat service
web2	192.168.0.105	Backend tomcat service
VIP	192.168.0.188	Virtual IP address

Installation service

Install on both master and backup servers

1, Install keepalived+nginx

1. Turn off firewall and enhance

systemctl stop firewalld.service
setenforce 0

2. Install the source with nginx rpm package

rpm -ivh http://nginx.org/packages/centos/7/noarch/RPMS/nginx-release-centos-7-0.el7.ngx.noarch.rpm

3. Install keepalived+nginx

yum install -y keepalived nginx

2, Install mysql

1. Check whether MySQL is installed in the system

rpm -qa | grep mysql

2. View the installed version of Mariadb database

rpm -qa|grep -i mariadb

3. Uninstall the installed Mariadb database.

rpm -qa|grep mariadb|xargs rpm -e --nodeps

4. Download the installation package file.

wget http://repo.mysql.com/mysql-community-release-el7-5.noarch.rpm

5. Install mysql-community-release-el7-5.noarch.rpm package

rpm -ivh mysql-community-release-el7-5.noarch.rpm

6. Install mysql.

yum install mysql-server -y

3, Install redis

yum install -y epel-release #To update epel source
yum install redis -y #Install redis

Install in node server (web1,web2)

4, Install tomcat

1. Install java environment

systemctl stop firewalld.service #Turn off firewall
setenforce 0
tar xf jdk-8u101-linux-x64.tar.gz -C /usr/local
cd /usr/local/
mv jdk1.8.0_101 java

2. Set environment variables

sed -i '$aJAVA_HOME=/usr/local/java\nCLASSPATH=$JAVA_HOME/lib\nPATH=$PATH:$JAVA_HOME/bin\nexport PATH JAVA_HOME CLASSPATH' /etc/profile
source /etc/profile
java -version

3. Install tomcat

tar xvf apache-tomcat-8.5.43.tar.gz -C /usr/local #decompression
cd /usr/local
mv apache-tomcat-8.5.43 tomcat #rename
chmod +x tomcat/bin/* #Authorize scripts
ln -s /usr/local/tomcat/bin/startup.sh /usr/bin/tomcatup #Establish a soft connection
ln -s /usr/local/tomcat/bin/shutdown.sh /usr/bin/tomcatdown

Change configuration and validation

1, Keep alive + nginx change configuration

1. Change the maintained configuration

vim /etc/keepalived/keepalived.conf #Modify three parameters

! Configuration File for keepalived

global_defs {
    route_id NGINX_HA   #The standby machine needs to be changed to nginx Gu HB
}

vrrp_script nginx {   #nginx monitoring script, which will be created later
    script "/opt/shell/nginx.sh"   
    interval 2
}

vrrp_instance VI_1 {
    state MASTER    #BACKUP instead of standby
    interface ens33
    virtual_router_id 53   #Two must be the same
    priority 100   #Priority, the standby machine is smaller than this, such as 90
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
}

track_script {
    nginx
}

virtual_ipaddress {
    192.168.0.188   #Virtual ip address
    }
}

2. Change nginx configuration

vim /etc/nginx/nginx.conf

 user  nginx;
worker_processes  1;

error_log  /var/log/nginx/error.log warn;
pid        /var/run/nginx.pid;

events {
    worker_connections  1024;
}

http {
    include       /etc/nginx/mime.types;
    default_type  application/octet-stream;

    log_format  main  '$remote_addr - $remote_user [$time_local] "$request" '
                      '$status $body_bytes_sent "$http_referer" '
                      '"$http_user_agent" "$http_x_forwarded_for"';

    access_log  /var/log/nginx/access.log  main;

    sendfile        on;
    #tcp_nopush     on;

    keepalive_timeout  65;

    #gzip  on;
upstream tomcatpool {
                                server 192.168.0.105:8080;  #ip address and port number of node server 1
                                server 192.168.0.104:8080; #ip address and port number of node server 2
                                ip_hash;   #Session stabilization function, otherwise you cannot log in through vip address
                }
                server {
                                listen 80;
                                server_name 192.168.0.188;  #Virtual ip address
                                location / {
                                                proxy_pass http://tomcatpool;
                                                proxy_set_header X-Real-IP $remote_addr;
                                }
    }
    include /etc/nginx/conf.d/*.conf;
}

3. Create nginx startup script

mkdir /opt/shell
vim /opt/shell/nginx.sh

#!/bin/bash
k=`ps -ef | grep keepalived | grep -v grep | wc -l`
if [ $k -gt 0 ];then
            /bin/systemctl start nginx.service
else
        /bin/systemctl stop nginx.service
 fi

chmod +x /opt/shell/nginx.sh #Script authorization
Remarks:2.platform nginx Service configuration, nginx Scripts are the same!

Verify the maintained + nginx function

1. View the ip messages of the active and standby

2. Stop the host keeepalive

Drainage process

Pits 1:

Solution: look at the picture information carefully, give a prompt, some naming error

Modify the nginx configuration file, change the tomcat_pool to Tomcat pool, and remember that both of them need to be modified, and restart the nginx service

Error cause analysis: This is because the new version of Nginx does not support this symbol when naming. Try to avoid using

Pit two:

Solution: check all configuration files carefully and make sure there is no problem. It should be a firewall problem. The firewalls of two servers must be closed or firewall policies must be made, otherwise VIP will not appear

Pit 3: after configuration, VIP (virtual ip) appears in both sets

Cause analysis:

Solution: modify the configuration file against the reason

Pit four:

Cause analysis: there is still a problem with the maintained configuration file

Solution: compare the above configuration file and modify it carefully to solve the problem. There must be something wrong with the details in the configuration file!

Database master slave synchronization

1. Change the configuration on the main database

binlog-ignore-db=mysql,information_schema
character_set_server=utf8
log_bin=mysql_bin
server_id=1
log_slave_updates=true
sync_binlog=1

2. Database authorization

3. Change the configuration from the database

server_id=2
relay-log=relay-log-bin
relay-log-index=slave-relay-bin.index

4. Synchronize settings from database

Project deployment of shopping mall

1. Decompress the project package

tar xf SLSaleSystem.tar.gz -C /usr/local/tomcat/webapps/

2. Modify tomcat configuration

#The function is to allow users to access directly without using the project name
<Context path="" docBase="SLSaleSystem" reloadable="true" debug="0"></Context>

3. Modify data link

cd /usr/local/tomcat/webapps/SLSaleSystem/WEB-INF/classes
vim jdbc.properties

driverClassName=com.mysql.jdbc.Driver
#Fill in virtual ip address
url=jdbc\:mysql\://192.168.0.188\:3306/slsaledb?useUnicode\=true&characterEncoding\=UTF-8
#Fill in the user name and password when authorizing
uname=root
password=123456
minIdle=10
maxIdle=50
initialSize=5
maxActive=100
maxWait=100
removeAbandonedTimeout=180
removeAbandoned=true

4. Start tomcat service

tomcatup
netstat -ntap |grep 8080

Note: the operation of two node servers is the same

5. Website verification

Redis master-slave and cluster configuration

On the host:

bind 0.0.0.0

systemctl start redis.service #Startup service
netstat -anpt | grep 6379
redis-cli -h 192.168.0.106 -p 6379 #Test connection
192.168.0.106:6379> set name test
OK
192.168.0.106:6379> get name
"test"
192.168.0.106:6379> exit

On the standby:

vi /etc/redis.conf

 bind 0.0.0.0  #61 elements
 slaveof 192.168.0.106  6379  #266 lines, IP of the primary server

redis is configured on the node server

#Fill in the virtual IP address, line 47
<constructor-arg value="192.168.0.188"/>
 <constructor-arg value="6379"/>

tomcatdown
tomcatup

redis cache test

redis-cli -h 192.168.0.188 -p 6379
192.168.0.188:6379> info

Configure redis cluster and master-slave switch

vim /etc/redis-sentinel.conf

protected-mode no  #17 elements
sentinel monitor mymaster 192.168.200.128 6379 1 #69 lines, 1 for 1 standby machine
sentinel down-after-milliseconds mymaster 3000 #98 lines, failover time in milliseconds

Service redis sentinel start
netstat -anpt | grep 26379

redis-cli -h 192.168.0.106 -p 26379 info Sentinel #View cluster information

Verify master-slave switch

Verify high availability architecture