运维监控-Open-Falcon安装Agent实战篇
作者：尹正杰
版权声明：原创作品，谢绝转载！否则将追究法律责任。
本文参考链接来自：http://book.open-falcon.org/zh/install_from_src/agent.html。
一.介绍
agent用于采集机器负载监控指标，比如cpu.idle、load.1min、disk.io.util等等，每隔60秒push给Transfer。agent与Transfer建立了长连接，数据发送速度比较快，agent提供了一个http接口/v1/push用于接收用户手工push的一些数据，然后通过长连接迅速转发给Transfer。
二.部署
agent需要部署到所有要被监控的机器上，比如公司有10万台机器，那就要部署10万个agent。agent本身资源消耗很少，不用担心。
1>.部署环境
2>.拷贝agent工具包到需要安装的服务器上
[root@node101 ~]# cd /yinzhengjie/open-falcon/workspace/open-falcon
[root@node101 open-falcon]#
[root@node101 open-falcon]# ll
total
drwxrwxr-x. Aug agent
drwxrwxr-x. Aug aggregator
drwxrwxr-x. Aug alarm
drwxrwxr-x. Aug api
drwxr-xr-x. root root Dec : data
drwxrwxr-x. Aug gateway
drwxrwxr-x. Aug graph
drwxrwxr-x. Aug hbs
drwxrwxr-x. Aug judge
drwxrwxr-x. Aug nodata
-rwxrwxr-x. Aug open-falcon
lrwxrwxrwx. Aug plugins -> ./agent/plugins/
lrwxrwxrwx. Aug public -> ./agent/public/
drwxrwxr-x. Aug transfer
[root@node101 open-falcon]#
[root@node101 ~]# cd /yinzhengjie/open-falcon/workspace/open-falcon #进入到“node101.yinzhengjie.org.cn”的open-Falcon安装目录
[root@node101 open-falcon]# scp -r agent/ root@node102.yinzhengjie.org.cn:~
The authenticity of host 'node102.yinzhengjie.org.cn (172.30.1.102)' can't be established.
ECDSA key fingerprint is SHA256:aGk7Wv77uRVAElcihCH0Zu0sAPp6qEs2sSA1Zsj3o7g.
ECDSA key fingerprint is MD5:2b::b8:ad:f8::::::f1::e7:::fb.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'node102.yinzhengjie.org.cn,172.30.1.102' (ECDSA) to the list of known hosts.
root@node102.yinzhengjie.org.cn's password:
agent.log % 41KB .6KB/s :
scheduler.go % .5KB/s :
reader.go % .4KB/s :
plugins.go % .9KB/s :
odometer.js % .3KB/s :
jquery.dataTables.min.js % 69KB .9MB/s :
jquery.js % 93KB .3MB/s :
jquery-ui.min.js % 223KB .5MB/s :
dashboard.js % .4MB/s :
bootstrap.js % 47KB .2MB/s :
base.js % .5KB/s :
code.png % .9KB/s :
favicon.ico % 97KB .5MB/s :
font-awesome.min.css % 22KB .0MB/s :
font-awesome-ie7.min.css % 37KB .5MB/s :
font-awesome-ie7.css % 40KB .7MB/s :
font-awesome.css % 27KB .8MB/s :
fontawesome-webfont.eot % 37KB .0KB/s :
FontAwesome.otf % 60KB .9MB/s :
fontawesome-webfont.woff % 43KB .5MB/s :
fontawesome-webfont.svg % 193KB .5MB/s :
fontawesome-webfont.ttf % 77KB .9MB/s :
odometer.css % .8MB/s :
dashboard.css % .1KB/s :
bootstrap-responsive.min.css % .0KB/s :
OpenSans-Regular-webfont.woff % 83KB .1MB/s :
OpenSans-Regular-webfont.eot % 71KB .4MB/s :
OpenSans-Regular-webfont.svg % 266KB .5MB/s :
OpenSans-Regular-webfont.ttf % 160KB .9MB/s :
style.css % 23KB .8MB/s :
bootstrap.min.css % 88KB .3MB/s :
g.css % .5KB/s :
index.html % 15KB .1MB/s :
falcon-agent % 9849KB .1MB/s :
cfg.json % .8KB/s :
[root@node101 open-falcon]#
[root@node101 open-falcon]# scp -r agent/ root@node102.yinzhengjie.org.cn:~ #拷贝agent目录到需要监控的"node102.yinzhengjie.org.cn"服务器上
[root@node101 open-falcon]# scp -r agent/ root@node103.yinzhengjie.org.cn:~
The authenticity of host 'node103.yinzhengjie.org.cn (172.30.1.103)' can't be established.
ECDSA key fingerprint is SHA256:aGk7Wv77uRVAElcihCH0Zu0sAPp6qEs2sSA1Zsj3o7g.
ECDSA key fingerprint is MD5:2b::b8:ad:f8::::::f1::e7:::fb.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'node103.yinzhengjie.org.cn,172.30.1.103' (ECDSA) to the list of known hosts.
root@node103.yinzhengjie.org.cn's password:
agent.log % 47KB .1MB/s :
scheduler.go % .1MB/s :
reader.go % .7MB/s :
plugins.go % .5MB/s :
odometer.js % .0KB/s :
jquery.dataTables.min.js % 69KB .0MB/s :
jquery.js % 93KB .0MB/s :
jquery-ui.min.js % 223KB .4MB/s :
dashboard.js % .9MB/s :
bootstrap.js % 47KB .1MB/s :
base.js % .2MB/s :
code.png % .0KB/s :
favicon.ico % 97KB .9MB/s :
font-awesome.min.css % 22KB .8MB/s :
font-awesome-ie7.min.css % 37KB .2MB/s :
font-awesome-ie7.css % 40KB .6MB/s :
font-awesome.css % 27KB .6MB/s :
fontawesome-webfont.eot % 37KB .7MB/s :
FontAwesome.otf % 60KB .6MB/s :
fontawesome-webfont.woff % 43KB .8MB/s :
fontawesome-webfont.svg % 193KB .1MB/s :
fontawesome-webfont.ttf % 77KB .4MB/s :
odometer.css % .1MB/s :
dashboard.css % .3MB/s :
bootstrap-responsive.min.css % .9MB/s :
OpenSans-Regular-webfont.woff % 83KB .9MB/s :
OpenSans-Regular-webfont.eot % 71KB .4MB/s :
OpenSans-Regular-webfont.svg % 266KB .4MB/s :
OpenSans-Regular-webfont.ttf % 160KB .1MB/s :
style.css % 23KB .9MB/s :
bootstrap.min.css % 88KB .3MB/s :
g.css % .0MB/s :
index.html % 15KB .0MB/s :
falcon-agent % 9849KB .5MB/s :
cfg.json % .1KB/s :
[root@node101 open-falcon]#
[root@node101 open-falcon]# scp -r agent/ root@node103.yinzhengjie.org.cn:~ #拷贝agent目录到需要监控的"node103.yinzhengjie.org.cn"服务器上
[root@node101 open-falcon]# scp open-falcon root@node102.yinzhengjie.org.cn:~
root@node102.yinzhengjie.org.cn's password:
open-falcon % 3894KB .9MB/s :
[root@node101 open-falcon]#
[root@node101 open-falcon]# scp open-falcon root@node102.yinzhengjie.org.cn:~ #拷贝open-falcon脚本到需要监控的服务器上面
[root@node101 open-falcon]# scp open-falcon root@node103.yinzhengjie.org.cn:~
root@node103.yinzhengjie.org.cn's password:
open-falcon % 3894KB .7MB/s :
[root@node101 open-falcon]#
[root@node101 open-falcon]# scp open-falcon root@node103.yinzhengjie.org.cn:~ #拷贝open-falcon脚本到需要监控的服务器上面
3>.修改node102.yinzhengjie.org.cn的cfg.json配置文件并启动agent
[root@node102 ~]# ll
total
drwxrwxr-x. root root Dec : agent
-rwxr-xr-x. root root Dec : open-falcon
[root@node102 ~]#
[root@node102 ~]# cat agent/config/cfg.json
{
"debug": true,
"hostname": "node102.yinzhengjie.org.cn",
"ip": "",
"plugin": {
"enabled": false,
"dir": "./plugin",
"git": "https://github.com/open-falcon/plugin.git",
"logs": "./logs"
},
"heartbeat": {
"enabled": true,
"addr": "node101.yinzhengjie.org.cn:6030",
"interval": ,
"timeout":
},
"transfer": {
"enabled": true,
"addrs": [
"node101.yinzhengjie.org.cn:8433"
],
"interval": ,
"timeout":
},
"http": {
"enabled": true,
"listen": ":1988",
"backdoor": false
},
"collector": {
"ifacePrefix": ["eth", "em"],
"mountPoint": []
},
"default_tags": {
},
"ignore": {
"cpu.busy": true,
"df.bytes.free": true,
"df.bytes.total": true,
"df.bytes.used": true,
"df.bytes.used.percent": true,
"df.inodes.total": true,
"df.inodes.free": true,
"df.inodes.used": true,
"df.inodes.used.percent": true,
"mem.memtotal": true,
"mem.memused": true,
"mem.memused.percent": true,
"mem.memfree": true,
"mem.swaptotal": true,
"mem.swapused": true,
"mem.swapfree": true
}
}
[root@node102 ~]#
{
"debug": true, # 控制一些debug信息的输出，生产环境通常设置为false
"hostname": "", # agent采集了数据发给transfer，endpoint就设置为了hostname，默认通过`hostname`获取，如果配置中配置了hostname，就用配置中的
"ip": "", # agent与hbs心跳的时候会把自己的ip地址发给hbs，agent会自动探测本机ip，如果不想让agent自动探测，可以手工修改该配置
"plugin": {
"enabled": false, # 默认不开启插件机制
"dir": "./plugin", # 把放置插件脚本的git repo clone到这个目录
"git": "https://github.com/open-falcon/plugin.git", # 放置插件脚本的git repo地址
"logs": "./logs" # 插件执行的log，如果插件执行有问题，可以去这个目录看log
},
"heartbeat": {
"enabled": true, # 此处enabled要设置为true
"addr": "127.0.0.1:6030", # hbs的地址，端口是hbs的rpc端口
"interval": , # 心跳周期，单位是秒
"timeout": # 连接hbs的超时时间，单位是毫秒
},
"transfer": {
"enabled": true, # 此处enabled要设置为true
"addrs": [
"127.0.0.1:8433",
"127.0.0.1:8433"
], # transfer的地址，端口是transfer的rpc端口, 可以支持写多个transfer的地址，agent会保证HA
"interval": , # 采集周期，单位是秒，即agent一分钟采集一次数据发给transfer
"timeout": # 连接transfer的超时时间，单位是毫秒
},
"http": {
"enabled": true, # 是否要监听http端口
"listen": ":1988" # 如果监听的话，监听的地址
},
"collector": {
"ifacePrefix": ["eth", "em"] # 默认配置只会采集网卡名称前缀是eth、em的网卡流量，配置为空就会采集所有的，lo的也会采集。可以从/proc/net/dev看到各个网卡的流量信息
},
"ignore": { # 默认采集了200多个metric，可以通过ignore设置为不采集
"cpu.busy": true,
"mem.swapfree": true
}
}
配置文件必须叫cfg.json，可以基于cfg.example.json修改（想要了解上述配置参数说明戳我）
./open-falcon start agent 启动进程
./open-falcon stop agent 停止进程
./open-falcon monitor agent 查看日志
看var目录下的log是否正常，或者浏览器访问其1988端口。另外agent提供了一个--check参数，可以检查agent是否可以正常跑在当前机器上
./falcon-agent --check
4>.修改node103.yinzhengjie.org.cn的cfg.json配置文件并启动agent
[root@node103 ~]# ll
total
drwxrwxr-x. root root Dec : agent
-rwxr-xr-x. root root Dec : open-falcon
[root@node103 ~]#
[root@node103 ~]# cat agent/config/cfg.json
{
"debug": true,
"hostname": "node103.yinzhengjie.org.cn",
"ip": "",
"plugin": {
"enabled": false,
"dir": "./plugin",
"git": "https://github.com/open-falcon/plugin.git",
"logs": "./logs"
},
"heartbeat": {
"enabled": true,
"addr": "node101.yinzhengjie.org.cn:6030",
"interval": ,
"timeout":
},
"transfer": {
"enabled": true,
"addrs": [
"node101.yinzhengjie.org.cn:8433"
],
"interval": ,
"timeout":
},
"http": {
"enabled": true,
"listen": ":1988",
"backdoor": false
},
"collector": {
"ifacePrefix": ["eth", "em"],
"mountPoint": []
},
"default_tags": {
},
"ignore": {
"cpu.busy": true,
"df.bytes.free": true,
"df.bytes.total": true,
"df.bytes.used": true,
"df.bytes.used.percent": true,
"df.inodes.total": true,
"df.inodes.free": true,
"df.inodes.used": true,
"df.inodes.used.percent": true,
"mem.memtotal": true,
"mem.memused": true,
"mem.memused.percent": true,
"mem.memfree": true,
"mem.swaptotal": true,
"mem.swapused": true,
"mem.swapfree": true
}
}
[root@node103 ~]#
[root@node103 ~]# ./open-falcon start agent
[falcon-agent]
[root@node103 ~]# .
[root@node103 ~]# ./open-falcon start agent #在node103.yinzhengjie.org.cn节点上启动agent
[root@node103 ~]# ./open-falcon monitor agent
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
// :: var.go:: => <Total=> <Endpoint:node101.yinzhengjie.org.cn, Metric:ss.estab, Type:GAUGE, Tags:, Step:, Time:, Value:>
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
// :: var.go:: => <Total=> <Endpoint:node101.yinzhengjie.org.cn, Metric:rate, Type:GAUGE, Tags:name=pfc.push.cnt, Step:, Time:, Value:0.016666781240232065>
// :: var.go:: => <Total=> <Endpoint:node101.yinzhengjie.org.cn, Metric:rate, Type:GAUGE, Tags:name=pfc.push.cnt, Step:, Time:, Value:0.01666669039142266>
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
// :: cfg.go:: read config file: /root/agent/config/cfg.json successfully
// :: http.go:: listening :
// :: var.go:: => <Total=> <Endpoint:node101.yinzhengjie.org.cn, Metric:df.bytes.free.percent, Type:GAUGE, Tags:mount=/,fstype=xfs, Step:, Time:, Value:97.8937870429458>
// :: var.go:: => <Total=> <Endpoint:node101.yinzhengjie.org.cn, Metric:agent.alive, Type:GAUGE, Tags:, Step:, Time:, Value:>
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
// :: var.go:: => <Total=> <Endpoint:node101.yinzhengjie.org.cn, Metric:ss.estab, Type:GAUGE, Tags:, Step:, Time:, Value:>
// :: var.go:: <= <Total=, Invalid:, Latency=0ms, Message:ok>
[root@node103 ~]# ./open-falcon monitor agent #查看agent端的日志信息
三.在open-falcon中的前端界面查看是否有对应的主机列表
1>.刷新前端页面，会看到2台监控的主机信息，如下图所示：
2>.以Endpoint视角查看agent采集的监控项目
3>.点击后可查看到的视图信息如下：
