当前位置：网站首页>Case: using kept+haproxy to build a Web Cluster

Case: using kept+haproxy to build a Web Cluster

2022-07-26 02:37:00 【Are prawns delicious】

Catalog

* Case environment

* build Nginx1

* build Nginx2

* Compilation and installation Haproxy

* install keepalived

* Case test

*Haproxy journal

*Haproxy Parameter optimization of

        Haproxy It is a popular cluster scheduling tool , There are many similar cluster scheduling tools , Such as LVS and Nginx. By comparison ,LVS Best performance , But the construction is relatively complex ;Nginx Of upstream Module supports clustering , However, the health check function of cluster nodes is not strong , No performance Haproxy good .Haproxy The official website is http://www.haproxy.org/.

Case pre knowledge points

  1. HTTP request Request mode GET The way POST The way Return status code The normal status code is 2××、3×× The abnormal status code is 4××、5××

  2. Common scheduling algorithms for load balancing

    RR（Round Robin）： Polling scheduling ​ LC（Least Connections）： Minimum connections ​ SH（Source Hashing）： Source based access scheduling

Case environment

         The goal of the experiment ： Use haproxy build web to cluster around , Achieve load balancing and high availability , Use keepalived+haproxy Realize dual machine hot standby and load balancing .

         Case preparation ： Configure according to the following figure IP Address , The drift address is 192.168.1.100, Turn off firewall 、selinux、 build yum Warehouse .

build Nginx1

[[email protected] ~]# yum -y install pcre-devel zlib-devel
[[email protected] ~]# useradd -M -s /sbin/nologin nginx
[[email protected] ~]# eject
[[email protected] ~]# mount /dev/cdrom /media
mount: /dev/sr0 is write-protected, mounting read-only
[[email protected] ~]# tar zxf /media/nginx-1.12.0.tar.gz -C /usr/src
[[email protected] ~]# cd /usr/src/nginx-1.12.0/
[[email protected] nginx-1.12.0]# ./configure --prefix=/usr/local/nginx --user=nginx --group=nginx
[[email protected] nginx-1.12.0]# make && make install

The default information after installation is as follows .

Default installation directory ：/usr/local/nginx

The default log ：/usr/local/nginx/logs

Listening port ：80

Default web Catalog ：/usr/local/nginx/html

[[email protected] nginx-1.12.0]# cd /usr/local/nginx/html/
[[email protected] html]# echo 111111 > test.html                              // Create a test page 
[[email protected] html]# /usr/local/nginx/sbin/nginx                          // start-up Nginx
[[email protected] html]# netstat -anpt | grep nginx
tcp        0      0 0.0.0.0:80              0.0.0.0:*               LISTEN      6762/nginx: master  

After installation , Client access http://192.168.1.20/test.html test .

build Nginx2

The steps of compiling and installing are the same as Nginx1 identical , The difference lies in the test web page .

[[email protected] html]# echo 22222 > test.html

After installation , Client access http://192.168.1.30/test.html test .

Compilation and installation Haproxy

[[email protected] ~]# yum -y install pcre-devel bzip2-devel 
[[email protected] ~]# eject
[[email protected] ~]# mount /dev/cdrom /media/
mount: /dev/sr0 is write-protected, mounting read-only
[[email protected] ~]# tar zxf /media/haproxy-1.5.19.tar.gz -C /usr/src
[[email protected] ~]# cd /usr/src/haproxy-1.5.19/
[[email protected] haproxy-1.5.19]# make TARGET=linux26                    //64 Bit system 
[[email protected] haproxy-1.5.19]# make install

Haproxy Server configuration

  • Build profile 

[[email protected] haproxy-1.5.19]# mkdir /etc/haproxy                             // Create a directory of configuration files 
[[email protected] haproxy-1.5.19]# cp examples/haproxy.cfg /etc/haproxy/          // take haproxy.cfg Copy the file to the configuration file directory

  • Haproxy Configuration item introduction

        Haproxy The configuration file is usually divided into three parts , namely global、defaults and listen.global Configure for global ,defaults For default configuration ,listen Configure... For application components .

global The configuration items are described as follows ：

global
        log 127.0.0.1   local0                      // Configure logging ,local0 For log devices , It is stored in the system log by default 
        log 127.0.0.1   local1 notice               //notice For log level , Usually there are 24 A level 
        maxconn 4096                                // maximum connection 
        uid 99                                      // user uid
        gid 99                                      // user gid
defaults The configuration items are described as follows ：

defaults
        log     global                              // Define the log as global Log definition in configuration 
        mode    http                                // The model is http
        option  httplog                             // use http Log format log 
        retries 3                                   // Check node server failures , Continuous over 3 Secondary failure , The node is considered unavailable 
        redispatch                                  // When the server load is high , Automatically end connections that have been queued for a long time 
        maxconn 2000                                // maximum connection 
        contimeout      5000                        // Connection timeout 
        clitimeout      50000                       // Client timeout 
        srvtimeout      50000                       // Server timeout 
global The configuration items are described as follows ：

listen  appli4-backup 0.0.0.0:10004                                                 // Define a appli4-backup Application 
        option  httpchk /index.html                                                 // Check the... Of the server index.html file 
        option  persist                                                             // Force the request to be sent to already down Dropped servers 
        balance roundrobin                                                          // The load balancing scheduling algorithm uses the polling algorithm 
        server  inst1 192.168.114.56:80 check inter 2000 fall 3                     // Define online nodes 
        server  inst2 192.168.114.56:81 check inter 2000 fall 3 backup              // Define backup nodes

  • According to the current cluster design , take haproxy.conf The content of the configuration file is modified as follows .

[[email protected] ~]# vim /etc/haproxy/haproxy.cfg
# this config needs haproxy-1.1.28 or haproxy-1.2.1
​
global
        log 127.0.0.1   local0
        log 127.0.0.1   local1 notice
        #log loghost    local0 info
        maxconn 4096
        #chroot /usr/share/haproxy                                      // Comment out 
        uid 99
        gid 99
        daemon
        #debug
        #quiet
​
defaults
        log     global
        mode    http
        option  httplog
        option  dontlognull
        retries 3
        #redispatch                                                     // Comment out 
        maxconn 2000
        contimeout      5000
        clitimeout      50000
        srvtimeout      50000
listen  appli4-backup 0.0.0.0:80                                        // Copy listen Configuration item module content , Change the port number to 80
        option  httpchk GET /index.html                                 //http The request method is changed to GET
        balance roundrobin
        server  inst1 192.168.1.20:80 check inter 2000 fall 3           // Configuration of two node servers 
        server  inst2 192.168.1.30:80 check inter 2000 fall 3           // Delete all the following configurations in the configuration file
  • Create a self starting script 
[[email protected] ~]# cp /usr/src/haproxy-1.5.19/examples/haproxy.init /etc/init.d/haproxy
[[email protected] ~]# ln -s /usr/local/sbin/haproxy /usr/sbin/haproxy
[[email protected] ~]# chmod +x /etc/init.d/haproxy 
[[email protected] ~]# chkconfig --add /etc/init.d/haproxy 
[[email protected] ~]# /etc/init.d/haproxy start
Starting haproxy (via systemctl):                          [  OK  ]

install keepalived

  • install ipvsadm and keepalived package .

[[email protected] ~]# mount /dev/cdrom /media
mount: /dev/sr0 is write-protected, mounting read-only
[[email protected] ~]# yum -y install ipvsadm keepalived

  • To configure keepalived.cof

         do haproxy+keepalived to cluster around , You only need to configure and modify the hot standby instance , Specify the primary and standby schedulers , name ,VIP Address , Priority and drift address are enough .

[[email protected] ~]# vim /etc/keepalived/keepalived.conf 
! Configuration File for keepalived
​
global_defs {
......// Omit part 
    router_id R1                    // Name of main regulator 
}  
​
vrrp_instance VI_1 {
    state MASTER                    // Hot standby state 
    interface ens33                 //VIP Address physical interface 
    virtual_router_id 51
    priority 100                    // priority 
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.1.100               // Drift address 
    }
}
[[email protected] ~]# systemctl start keepalived
[[email protected] ~]# systemctl enable keepalived

        Haproxy+keepalived The installation configuration of the backup scheduler is roughly the same , The only thing to notice is that keepalived When the configuration , Distinguish the main 、 Standby scheduler configuration （ Specify the primary and standby schedulers , name ,VIP Address , Priority and drift address ）. Continue to configure according to the above configuration haproxy2+keepalived2.

Case test

  • The client opens the browser to access http://192.168.1.100/test.html, Refresh the browser twice to view the web content and test the load balance （ You should see 11111 and 22222 Two web pages ）. To break off web1, Refresh twice in access to test high availability （ At this time, only 22222）.

  • broken keepalived Master tuner network , Continue to visit http://192.168.1.100/test.html, Successful access means keepalived Load balancing successful （ You should see 11111 and 22222 Two web pages ）.

Haproxy journal

         Modify the log configuration options in the original configuration file , stay global Configuration in the project . Save and restart after modifying the configuration Haproxy.

[[email protected] ~]# vim /etc/haproxy/haproxy.cfg
        log /dev/log    local0 info
        log /dev/log    local0 notice
[[email protected] ~]# /etc/init.d/haproxy restart             // restart 
Restarting haproxy (via systemctl):                        [  OK  ]

         For ease of management , take Haproxy The relevant configuration is defined independently to haproxy.conf And on the /etc/rsyslog.d/ Next ,rsyslog All configuration files in this directory will be loaded automatically at startup .

[[email protected] ~]# vim /etc/rsyslog.d/haproxy.conf
if ($programname == 'haproxy' and $syslogseverity-text == 'info') then -/var/log/haproxy/haproxy-info.log
& ~ 
if ($programname == 'haproxy' and $syslogseverity-text == 'notice') then -/var/log/haproxy/haproxy-notice.log
& ~
[[email protected] ~]# systemctl restart rsyslog

After client access , You can use the following command to check in time Haproxy Access request log information for .

[[email protected] ~]# tail -f /var/log/haproxy/haproxy-info.log
Jul 23 23:08:35 localhost haproxy[43674]: 192.168.1.200:62207 [23/Jul/2022:23:08:35.157] appli4-backup appli4-backup/inst2 0/0/1/1/2 200 240 - - ---- 1/1/0/1/0 0/0 "GET /test.html HTTP/1.1"
Jul 23 23:08:35 localhost haproxy[43674]: 192.168.1.200:62207 [23/Jul/2022:23:08:35.159] appli4-backup appli4-backup/inst1 198/0/5/1/204 200 241 - - ---- 2/2/0/1/0 0/0 "GET /test.html HTTP/1.1"

Haproxy Parameter optimization of

Parameters Parameter description Optimization Suggestions
maxconn maximum connection This parameter is adjusted according to the actual use of the application , Recommended 10240, meanwhile "defaults" The value of the maximum number of connections in cannot exceed "global" The definition in paragraph
daemon Daemons mode Haproxy You can use non - daemons mode to start , It is recommended to use daemon mode to start the production environment
nbproc Number of concurrent processes for load balancing It is recommended to contact the current server CPU Equal or equal in number of cores 2 times
retries Retry count Parameters are mainly used to check cluster nodes , If there are many nodes , And the volume is large , Set to 2 Time or 3 Time ; When there are few server nodes , You can set 5 Time or 6 Time
option http-server-close Active shut down http Request options It is recommended to use this option in the production environment , Avoid due to timeout Caused by too long time setting http Connection stacking
timeout http-keep-alive Long timeout This option sets the long connection timeout , For details, refer to the application's own characteristics , It can be set to 10s
timeout http-requestHTTP Request timeout It is recommended that this time be set to 5~10s, increase http Connection release speed
timeout client Client timeout If the traffic is too large , Slow node response , You can set this time to be shorter , Recommended setting is 1min Left and right
原网站

版权声明
本文为[Are prawns delicious]所创，转载请带上原文链接，感谢
https://yzsam.com/2022/207/202207260231271307.html

边栏推荐

猜你喜欢

随机推荐