varnish 使用手册
一、安装varnish
1、编译安装
- 安装依赖环境包
yum -y install autoconf automake jemalloc-devel libedit-devel libtool ncurses-devel pcre-devel pkgconfig python-docutils python-sphinx
- 安装varnish
wget http://varnish-cache.org/_downloads/varnish-6.0.0.tgz
tar xf varnish-6.0.0.tgz
cd varnish-6.0.0
sh autogen.sh
sh configure
make
make install
2、rpm安装
- 配置yum 源
vim /etc/yum.repo.d/varnish.repo
[varnishcache_varnish41]
name=varnishcache_varnish41
baseurl=https://packagecloud.io/varnishcache/varnish41/el/6/$basearch
repo_gpgcheck=1
gpgcheck=0
enabled=1
gpgkey=https://packagecloud.io/varnishcache/varnish41/gpgkey
sslverify=1
sslcacert=/etc/pki/tls/certs/ca-bundle.crt
metadata_expire=300
[varnishcache_varnish41-source]
name=varnishcache_varnish41-source
baseurl=https://packagecloud.io/varnishcache/varnish41/el/6/SRPMS
repo_gpgcheck=1
gpgcheck=0
enabled=1
gpgkey=https://packagecloud.io/varnishcache/varnish41/gpgkey
sslverify=1
sslcacert=/etc/pki/tls/certs/ca-bundle.crt
metadata_expire=300
- 安装varnish
yum -y install varnish jemalloc-devel
二、varnish 工作流程
1)varnish从客户端接收请求后，由vcl_recv状态引擎处理，不能识别的请求将会通过参数pipe交给vcl_pipe状态引擎，需要查找缓存的
请求通过lookup参数将会交给vcl_hash状态引擎，无需缓存的数据通过参数pass将会交给 vcl_pass状态引擎；
2)vcl_hash状态引擎在接收到请求后会从缓存中查找数据，查询结果有两种，一种是hit缓存命中，另一种是miss缓存未命中；
3)vcl_hit状态引擎将命中的缓存数据通过参数deliver交给vcl_deliver状态引擎，vcl_deliver状态引擎将数据处理后，最终返回给客户端；
4)vcl_miss状态引擎将未命中的结果参数fetch交给vcl_fetch状态引擎，vcl_fetch状态引擎将会从数据库中查找数据；
5)vcl_fetch状态引擎将从数据库中查询到的结果，返回给vcl_deliver状态引擎；
6)vcl_deliver状态引擎将结果返回给master进程，最终返回给客户端；
三、varnish 配置
1、/etc/sysconfig/varnish
#最大文件数
NFILES=131072
#共享锁内存
MEMLOCK=82000
#最大线程数
NPROCS="unlimited"
#
RELOAD_VCL=1
#varnish 默认规则配置文件
VARNISH_VCL_CONF=/etc/varnish/default.vcl
#varnish 监听的端口
VARNISH_LISTEN_PORT=80
#varnish 监听的IP
VARNISH_ADMIN_LISTEN_ADDRESS=127.0.0.1
#varnish 管理的端口
VARNISH_ADMIN_LISTEN_PORT=6082
#varnish 管理秘钥
VARNISH_SECRET_FILE=/etc/varnish/secret
#varnish 最小线程数
VARNISH_MIN_THREADS=50
#varnish 最大线程数
VARNISH_MAX_THREADS=1000
#varnish 内存块大小
VARNISH_STORAGE_SIZE=256M
#存储模型
VARNISH_STORAGE="malloc,${VARNISH_STORAGE_SIZE}"
#后端TTL
VARNISH_TTL=120
#启动参数配置
DAEMON_OPTS="-a ${VARNISH_LISTEN_ADDRESS}:${VARNISH_LISTEN_PORT} \
-f ${VARNISH_VCL_CONF} \
-T ${VARNISH_ADMIN_LISTEN_ADDRESS}:${VARNISH_ADMIN_LISTEN_PORT} \
-p thread_pool_min=${VARNISH_MIN_THREADS} \
-p thread_pool_max=${VARNISH_MAX_THREADS} \
-S ${VARNISH_SECRET_FILE} \
-s ${VARNISH_STORAGE}"
2、VCL内置函数
vcl_backend_error 当从后端主机获取源文件失败后，调用此函数
vcl_backend_fetch 向后端主机发送请求钱，调用此函数，可修改发往后端的请求
vcl_backend_response 获得后端主机的响应后，调用此函数
vcl_deliver 将在缓存中找到请求的内容发送给客户端前调用此方法
vcl_fini 当所有请求都离开当前VCL，且当前VCL被弃用时，调用此函数，经常用于清理varnish模块
vcl_hash 在vcl_recv调用后为请求创建一个hash值，调用此函数，此hash值将作为varnish中搜索缓存对象的key
vcl_hit 在执行lookup指令后，在缓存中找到请求的内容将自动调用该函数
vcl_init VCL加载时调用此函数，经常用于初始化varnish模块（VMODs）
vcl_miss 在执行lookup指令后，在缓存中没有找到请求的内容时自动调用该函数，此函数可用于判断是否需要从后端服务器获取内容
vcl_pass 此函数在进入pass模式时被调用，用于将请求直接传递至后端主机，并将后端响应原样返回客户端
vcl_pipe 此函数在进入pipe模式时被调用，用于将请求直接传递至后端主机，但后端主机的响应并不缓存直接返回客户端
vcl_purge pruge操作执行后调用此函数，可用于构建一个响应
vcl_recv 用于接收和处理请求，当请求达到并成功接收后被调用，通过判断请求的数据来决定如何处理请求
3、backend 配置
默认配置
backend default {
.host = "127.0.0.1";
.port = "8080";
}
健康检查
#直接配置：
backend default {
.host = "127.0.0.1";
.port = "8080";
.probe = {
.url = "/";
.timeout = 1s;
.interval = 5s;
.window = 5;
.threshold = 3;
}
}
健康检查页面是 "/" ,超时时间为1秒，每5秒检查一次，5次有3次成功认为是健康的
#以模块配置：
probe backend_healthcheck {
.url = health.html"
.windows = 5;
.threshold = 2;
.interval = 3s;
}
backend web1 {
.host = "127.0.0.1";
.port = "8080";
.probe = backend_healthcheck;
}
backend web2 {
.host = "127.0.0.1";
.port = "8080";
.probe = backend_healthcheck;
}
重新定义
backend web {
.host = "127.0.0.1";
.port = "8080";
}
vcl_recv {
set req.backend_hint = web ;
}
多主机
#基于权重：
import directors
probe backend_healthcheck {
.url = health.html"
.windows = 5;
.threshold = 2;
.interval = 3;
}
backend web1 {
.host = "192.168.1.1";
.port = "80";
.probe = backend_healthcheck;
}
backend web2 {
.host = "192.168.1.2";
.port = "80";
.probe = backend_healthcheck;
}
sub vcl_init {
new web_cluster = directors.random();
web_cluster.add_backend(web1,1.0);
web_cluster.add_backend(web2,1.0);
}
sub vcl_recv {
set req.backend_hint = web_cluster;
}
#基于请求类型：
backend html {
.host = "192.168.1.1";
}
backend php {
.host = "192.168.1.2";
}
sub vcl_recv {
if ( req.url ~ "\.php" ){
set req.backend_hint = php;
}
set req.backend_hint = html;
}
#基于域名：
backend moblie {
.host = "192.168.1.1";
}
backend host {
.host = "192.168.1.2";
}
backend default {
.host = "192.168.1.3";
}
sub vcl_recv {
if ( req.http.host == "moblie.demo.com" || req.http.host == "m.demo.com" ){
set req.backend_hint = moblie;
} elsif ( req.http.host == "www.demo.com" || req.http.host == "demo.com" ) {
set req.backend_hint = host;
}
set req.backend_hint = default;
}
4、 添加缓存标识
sub vcl_deliver {
if ( obj.hits > 0 ) {
set resp.http.X-cache = "HIT from " + server.ip;
} else {
set resp.http.X-cache = "MISS " + server.ip;
}
}
5、缓存清理
#定义配置
acl purge_ip {
"localhost";
"127.0.0.1";
}
sub vcl_recv {
if ( req.method ~ "PURGE" ){
if ( client.ip ~ purge_ip ) {
return(purge);
}
return(synth(404,"Not Found"));
}
}
sub vcl_purge {
return(synth(200,"Success"));
}
#清除缓存
curl -X PURGE http://www.demo.com/test.html
#规则清理
varnishadm
>ban req.url /test.html
6、Grace mode （优雅模式）
#过期仍然存放2分钟
sub vcl_backend_response {
set beresp.grace = 2m;
}
#如果健康检查成功，直接返回。健康检查失败，执行grace 模式
import std;
if (!std.healthy(req.backend_hint) && (obj.ttl + obj.grace > 0s)) {
return (deliver);
} else {
return (fetch);
}
7、清除cookie
sub vcl_recv {
if (req.url ~ "^/images") {
unset req.http.cookie;
}
}
8、改变后端响应
sub vcl_backend_response {
if (bereq.url ~ "\.(png|gif|jpg)$") {
unset beresp.http.set-cookie;
set beresp.ttl = 1h;
}
}
9、配置ACL
#只允许 localhost ，192.168.1.0/24 清除缓存
acl local {
"localhost";
"192.168.1.0"/24; /* and everyone on the local network */
! "192.168.1.23"; /* except for the dialin router */
}
sub vcl_recv {
if (req.method == "PURGE") {
if (client.ip ~ local) {
return(purge);
} else {
return(synth(403, "Access denied."));
}
}
}
#禁止 192.168.1.0/24 访问
acl local {
"192.168.1.0"/24; /* and everyone on the local network */
}
sub vcl_recv {
if (client.ip ~ local) {
return(synth(403, "Access denied."));
}
}
10、添加头信息
if (req.restarts == 0) {
if (req.http.x-forwarded-for) {
set req.http.X-Forwarded-For =
req.http.X-Forwarded-For + ", " + server.ip;
} else {
set req.http.X-Forwarded-For = server.ip;
}
}
11、定义缓存策略
sub vcl_recv {
if (req.url ~ "\.php$") {
return(pass);
}
}
sub vcl_fetch {
if (bereq.url ~ "\.(jpg|jpeg|gif|png)$") {
set beresp.ttl = 7200s;
}
if (bereq.url ~ "\.(html|css|js)$") {
set beresp.ttl = 1200s;
}
}
12、清除头信息
sub vcl_deliver {
unset resp.http.Via;
unset resp.http.X-Varnish;
}
四、命令工具
- varnishadm
- varnishhist
- varnishncsa
- varnishstat
- varnishtop
- varnishlog
- varnish_reload_vcl
- varnishtest