keepalived高可用haproxy配合varnish实现wordpress的动静分离

来源:互联网 发布:linux的安全审计服务 编辑:程序博客网 时间:2024/05/29 07:13

haproxy和nginx都可以作为七层和四层反代服务器对外提供服务,此文通过haproxy和keealived配置varnish搭建wordpress的动静分离站点

一、实验环境

五台虚拟机

  • haproxy-1:搭建haproxy和keepalived服务,ip地址:192.168.11.176

  • haproxy-2:搭建haproxy和keepalived服务,ip地址:192.168.11.172

  • varnish服务器ip地址:192.168.11.173

  • wordpress静态服务器:搭建httpd,ip地址:192.168.11.177

  • wordpress动态服务器,搭建httpd+php+mariadb,ip地址:192.168.11.178

实验目的

通过haproxy访问到后台网站,并且varnish为网站提供缓存,提升网站性能。

实验拓扑:此处虚拟机数量不够,就不采用动静分离集群了,而是分别采用一台虚拟机部署:

二、实验步骤

wordpress配置

  • 动态服务器配置:

    1. 安装httpd,mariadb,php服务,此处php作为httpd的模块进行动态php文件处理:

      yum install -y httpd mariadb-server php php-mysql php-mbstring
    2. 上传wordpress到/var/www/html路径下并修改属主属组为apache,此处不做详细演示;

  • 静态服务器配置:

    1. 安装httpd服务并上传wordpress到/var/www/html,修改属主属组为apache;

varnish配置

  1. 安装varnish服务并修改varnish的配置文件default.vcl,修改内容如下:

    vcl 4.0;      #声明为varnish4.0import directors;   #调用多个后端主机做集群#probe health_check {    #定义健康状态检测机制#    .url = "/.health_check.html";   #定义检测文件#    .window = 5;    #定义一共检测几次#    .threshold = 3; #定义检测几次失败则认为后端主机失效#    .interval = 2s; #每隔2秒检测一次#    .timeout = 3s;  #连接超时时长为3秒#}backend dynamic {   #定义后端动态主机#    .host = "192.168.11.178";    .port = "80";    .probe = health_check;}backend static {    #定义前端主机#    .host = "192.168.11.177";    .port = "80";    .probe = health_check;}sub vcl_recv {    # Happens before we check if we have this in cache already.    #    # Typically you clean up the request here, removing cookies you don't need,    # rewriting the request, etc.    if (req.url ~ "(?i).*php.*") {      #当访问的url包括php,则转到动态服务器,否则所有请求转到静态服务器,实现动静分离#        set req.backend_hint = dynamic;    } else {        set req.backend_hint = static;    }    if (req.url ~ "(?i)(register|login)$") {    #如果请求地址结尾为register或者login,注册和登录页面,则不查缓存,,直接从后端服务器获取内容#        return (pass);    }}sub vcl_backend_response {    # Happens after we have read the response headers from the backend.    #    # Here you clean the response headers, removing silly Set-Cookie headers    # and other mistakes your backend does.    #当后端服务器回复给varnish的响应如果不包含公共缓存信息,而且请求为jpg等静态资源,则卸载cookie信息并缓存资源1小时#    if (beresp.http.cache-control !~ "s-maxage") {          if (bereq.url ~ "(?i)\.(jpg|jpeg|png|gif|css|js|xml)$") {                unset beresp.http.Set-Cookie;                set beresp.ttl = 3600s;        }    }      #当varnish请求后端服务器的url包括php,则卸载cookie信息并缓存资源1小时#    if (bereq.url ~ "(?i).*php.*") {        unset beresp.http.Set-Cookie;        set beresp.ttl = 3600s;    }}sub vcl_deliver {    # Happens when we have all the pieces we need, and are about to send the    # response to the client.    #    # You can do accounting or modifying the final object here.    if (obj.hits>0) {        set resp.http.X-Cache = "HIT via "+server.ip;    } else {        set resp.http.X-Cache = "MISS via "+server.ip;    }}

    注意:

    varnish在探测到请求和响应报文头部有cookie信息的时候是不缓存的,所以缓存命中率会非常低。这就是为什么要卸载php页面和jpg等动态资源cookie的原因。

  2. 然后通过varnishadm登录varnish并加载配置项,用ab进行压测会发现启用和不启用缓存的性能差出数倍,而且启用和不启用缓存后端服务器压力也差出数倍;

haproxy配置

  1. 修改/etc/haproxy/haproxy.cfg配置文件,内容如下:

    global        #全局配置#    log         127.0.0.1 local2        #log记录到rsyslog服务器,此处需要在rsyslog进行配置#    chroot      /var/lib/haproxy        #把haproxy禁锢到/var/lib/haproxy,防止haproxy被劫持后破坏系统    pidfile     /var/run/haproxy.pid    #pid文件位置#    maxconn     4000                    #每个进程最大连接数为4000#    user        haproxy                     group       haproxy    daemon                              #以服务方式运行#    stats socket /var/lib/haproxy/stats    defaults    #默认配置#    mode                    http    #默认为http七层代理#    log                     global  #日志采用global配置#    option                  httplog #以http方式记录日志#    option                  dontlognull #不记录健康状态检测日志#    option http-server-close        option forwardfor       except 127.0.0.0/8  #传递客户端ip到后端服务器,需要在后端服务器修改日志样式#    option                  redispatch    retries                 3   #最大尝试3次连接#    timeout http-request    10s #等待http请求时间为10s#    timeout queue           1m  #    timeout connect         10s #等待建立连接时间为10s#    timeout client          1m  #等待客户端连接时间#    timeout server          1m  #等待服务端连接时间#    timeout http-keep-alive 10s #长连接时长#    timeout check           10s #检测间隔#    maxconn                 3000        listen stats    #定义状态页#    bind        *:9000    stats       enable    stats auth  admin:admin    stats uri   /admin?stats    #定义stats url路径#    stats refresh       30s    stats realm         "status-page"    stats hide-version  #隐藏版本信息#    stats admin if TRUE #开启后端管理功能#    frontend  web   #定义前端服务器名为web#    bind        *:80    default_backend     appsrvs #绑定的后端服务器#    backend appsrvs #定义后端服务器名为appsrvs#    server      static  192.168.11.173:80 check #定义后端服务器,static为名字,check表示进行健康状态检测,也可以添加其他值进来,如检测间隔等信息#

  2. 状态页面如下:

keepalived配置

  1. 修改/etc/keepalived/keepalived.conf配置文件,内容如下:

    ! Configuration File for keepalivedglobal_defs {   notification_email {     acassen@firewall.loc     failover@firewall.loc     sysadmin@firewall.loc   }   notification_email_from Alexandre.Cassen@firewall.loc   smtp_server 192.168.200.1   smtp_connect_timeout 30   router_id node1   vrrp_mcast_group4 224.0.115.15}vrrp_script chk_haproxy {    script "pkill -0 haproxy && exit 0 || exit 1"    interval 1    weight -5    fall 2    rise 1}vrrp_instance VI_1 {    state BACKUP    interface ens33    virtual_router_id 10    priority 96    advert_int 1    authentication {        auth_type PASS        auth_pass hello    }    virtual_ipaddress {        192.168.11.200 label ens33:0    }    track_script {        chk_haproxy    }    notify_master "/etc/keepalived/notify.sh master"    notify_backup "/etc/keepalived/notify.sh backup"    notify_fault "/etc/keepalived/notify.sh fault"}

  2. 写notify脚本,内容如下:

     
    #!/bin/bash#contact='root@localhost'notify() {        local mailsubject="$(hostname) to be $1,vip floating."        local mailbody="$(date + '%F %T'):vrrp transition,$(hostname) changed to be $1."        echo "$mailbody" | mail -s "$mailsubject" $contact}case $1 inmaster)        notify master;;backup)        notify backup;;fault)        notify fault;;*)        echo "Usage: $(basename $0) {master|backup|fault}"        exit 1;;esac
  3. 定义的haproxy的虚拟ip为192.168.11.200,此时任何一个haproxy主机或者haproxy服务故障都不会影响网站的正常访问,通过192.168.11.200即可访问到网站。

此时实验配置成功,而且网站通过varnish可承载大量并发,但是仍有问题,前台haproxy为高可用,但是后台varnish和动静服务器均是单点,为单点故障,所以还需要进一步完善。

网站访问流程为:client –> 前端MASTER haproxy –> varnish –> 后端动静服务器。

阅读全文
0 0
原创粉丝点击