puppet的基础环境介绍：
puppet服务器和客户端都已下载了epel的外部yum源，都已通过yum程序自动安装了puppet程序，过程比较简单，这里就不一一介绍了，机器都放置在同一局域网内，cn7788.com的域名，内部有内网DNS环境，没有用LDP作为域控，局域网还有其它客户端，由于不需要使用puppet环境，所以就不一一介绍了。
server.cn7788.com 192.168.1.124 puppet-master client.cn7788.com 192.168.1.125 puppet-client lamp.cn7788.com 192.168.1.126 puppet-client xen.cn7788.com 192.168.1.144 puppet-client
大家可以将上面的域名对应关系可将其都写在各自机器的/etc/hosts文件里，在各个puppet客户端上建议ntpdate精准对时（因为puppet的证书对时间要求严格），不然puppet-client连接时会报如下错误：
warning: peer certificate won't be verified in thisSSL session info: Caching certificate for client.cn7788.com info: Caching certificate_revocation_list for ca err: Could not retrieve catalog from remote server:certificate verify failed. This is oftenbecause the time is out of sync on the server or client warning: Not using cache on failed catalog err: Could not retrieve catalog; skipping run err: Could not send report: certificate verifyfailed. This is often because the timeis out of sync on the server or client
需求如下：客户机机器xen.cn7788.com和lamp.cn7788.com没有安装nagios客户端程序，这时想过通过puppet-server推送SHELL脚本自动安装，其它的客户端暂时没这么需求，这个应该如何实现呢？
由于客户端节点机器比较多，所以这里需要用到节点和模块的概念，这里我们先建立名为nagioscli的模块，如下所示：
mkdir -p/etc/puppet/modules/nagioscli/{manifests,files,templates}
files目录下的nagioscli.sh文件内容如下所示：
#!/bin/bash useradd nagios cd /usr/local/src wget wget http://syslab.comsenz.com/downloads/linux/nagios-plugins-1.4.13.tar.gz wget http://syslab.comsenz.com/downloads/linux/nrpe-2.12.tar.gz tar zxvf nagios-plugins-1.4.13.tar.gz cd nagios-plugins-1.4.13 ./configure make make install chown nagios:nagios /usr/local/nagios chown -R nagios:nagios /usr/local/nagios/libexec cd ../ tar zxvf nrpe-2.12.tar.gz cd nrpe-2.12 ./configure make all make install-plugin make install-daemon make install-daemon-config sed -i's@allowed_hosts=127.0.0.1@allowed_hosts=114.112.11.11@'/usr/local/nagios/etc/nrpe.cfg #114.112.11.11为nagios服务器的IP地址，这个可以根据实际需求更改。 /usr/local/nagios/bin/nrpe -c/usr/local/nagios/etc/nrpe.cfg -d echo "/usr/local/nagios/bin/nrpe -c/usr/local/nagios/etc/nrpe.cfg -d" >> /etc/rc.local
site.pp文件内容如下：
import "node.pp"
这里扩展了site.pp文件内容，它会载入node.pp文件，这样puppet-master在启动的时候，就会自动截入并处理node.pp文件了。
node.pp文件内容如下所示：
node 'lamp.cn7788.com'{
file
{"/usr/local/src/nagioscli.sh":
source =>"puppet://server.cn7788.com/modules/nagioscli/nagioscli.sh",
group => root,
owner => root,
mode => 755,
}
exec {
"auto install naigios client":
command =>"sh /usr/local/src/nagioscli.sh",
user =>"root",
path =>["/usr/bin","/usr/sbin","/bin","/bin/sh"],
}
}
node 'xen.cn7788.com'{
file
{"/usr/local/src/nagioscli.sh":
source =>"puppet://server.cn7788.com/modules/nagioscli/nagioscli.sh ",
group => root,
owner => root,
mode =>644,
}
exec {
"auto install naigios client":
command =>"sh /usr/local/src/nagioscli.sh",
user =>"root",
path =>["/usr/bin","/usr/sbin","/bin","/bin/sh"],
}
}
node 'client.cn7788.com'{
｝
client.cn7788.com节点机器后面什么都没有，则表示没有任何操作在此节点机器上面，因为client机器也在puppet环境里，并配置成了自动连接，配置成如此，是防止自动连接时puppet频繁报错。
这里以xen.cn7788.com为例，在其主机上输入如下命令：
puppetd --test --server server.cn7788.com
xen.cn7788.com上命令显示结果如下所示：
info: Caching catalog for xen.cn7788.com
info: Applying configuration version '1382622383'
--- /usr/local/src/nagioscli.sh 2013-10-24 22:35:36.000000000 +0800
+++ /tmp/puppet-file.22857.0 2013-10-24 22:39:08.000000000 +0800
@@ -1,4 +1,5 @@
#!/bin/bash
+yum -y install httpd gcc gcc-c++ glibcglibc-common gd gd-devel
useraddnagios
cd/usr/local/src
wgetwget http://syslab.comsenz.com/downloads/linux/nagios-plugins-1.4.13.tar.gz
info: FileBucket adding{md5}f75e9aa3fc301c8e9c85f2677feaa9b5
info:/Stage[main]//Node[xen.cn7788.com]/File[/usr/local/src/nagioscli.sh]:Filebucketed /usr/local/src/nagioscli.sh to puppet with sumf75e9aa3fc301c8e9c85f2677feaa9b5
notice:/Stage[main]//Node[xen.cn7788.com]/File[/usr/local/src/nagioscli.sh]/content: contentchanged '{md5}f75e9aa3fc301c8e9c85f2677feaa9b5' to'{md5}a1ed4dc2b98450e3144530f32677f736'
notice:/Stage[main]//Node[xen.cn7788.com]/Exec[auto install naigios client]/returns:executed successfully
notice: Finished catalog run in 283.11 seconds
执行时间比较长，总共耗时283.11秒，我们要检查下xen.cn7788.com的节点机器上是否开启了nrpe 进程，输入命令如下所示：
ps aux | grep nrpe | grep –v grep
命令显示结果如下所示：
nagios 22331 0.0 0.1 5108 924 ? Ss 22:35 0:00/usr/local/nagios/bin/nrpe -c /usr/local/nagios/etc/nrpe.cfg -d
我们检查下/etc/rc.local，看此命令有没有添加进去，命令如下：
grep -v"^#" /etc/rc.local
命令执行结果显示如下所示：
touch /var/lock/subsys/local /usr/local/nagios/bin/nrpe -c/usr/local/nagios/etc/nrpe.cfg -d
检查结果说明puppet-master的nagioscli模块是正常的，lamp.cn7788.com的结果类似，这里就不再贴出检测结果了，我们主要看下lamp.cn7788.com总共耗时多少，命令如下所示：
puppetd --test --serverserver.cn7788.com
结果如下所示：
info: Caching catalog for lamp.cn7788.com info: Applying configuration version '1382622383' notice: /Stage[main]//Node[lamp.cn7788.com]/Exec[autoinstall naigios client]/returns: executed successfully notice: Finished catalog run in 169.08 seconds
执行时间比较长，总共耗时169.08秒。
其实工作中像这种推送脚本执行的需求还是很多的，类似在各种不同名字的节点上执行的优化服务器命令、批量清除varnish缓存加速服务器缓存、根据机器名推送文件，我们只需要将此案例稍为变通下即可在工作中投入应用了。