配置文件说明
本篇是说明icinga2的配置文件。跟nagios比较,逻辑思维是一样的,定义主机(组)、服务(组)、检测命令、模板、检测频率等。但是实际使用的语法却不一样,重新定义了一套关键词。具体细节可参考下文。有些地方我也没能搞明白,希望读者童鞋能一起探讨。
默认采用yum安装的icinga2.
1 matser server上配置文件的两个目录:
/etc/icinga2/, 更多的配置放在./conf.d下,这里主要是用来自定义配置。
文件名,只要你能明白是用来做什么的即可,不需要一定区分user,service什么的。
/usr/share/icinga2/include/,这里主要是一些已经定义好的命令,可以直接使用。同时也可以参考这些定义好的命令, 来实现自己的脚本插件.
2 各个配置文件的说明
commands.conf和command-plugins.conf 定义命令
object CheckCommand "ssh"{ import"plugin-check-command" command = PluginDir+"/check_ssh" #defined in constants.conf arguments = { "-p"= "$ssh_port$" "host"= { value = "$ssh_address$" skip_key = true order = -1 } } vars.ssh_address= "$address$" }
object CheckCommand 为定义命令的固件关键词
import导入模板command.conf里的
Command用法,PluginDir定义在constants.conf
Arguments参数,如果是自定义的脚本,可以不需要在这里定义命令
"-p"= "$ssh_port$" 这个-p是插件本身的参数,后面的ssh_port是自定义名,格式$.....$
templates.conf 定义模板
针对host的检测模板:
template Host "generic-host" { max_check_attempts = 3 check_interval = 5m retry_interval = 30s check_command = "hostalive" }
针对service的检测模板:
template Service "generic-service" { max_check_attempts = 2 check_interval = 5m retry_interval = 20s }
template Host templateService固定格式,后面引号内名字自定义
max_check_attempts检测遇到问题,最多尝试次数
check_interval 检测的频率
retry_interval 如果检测遇到问题,重新检测的频率
通知模板:
template Notification "30mins-notification" { interval = 30m command = "mail-service-notification" states = [ Critical ] types = [ Problem, Recovery ] period = "24x7" }
Command定义在commands.conf里
States这里设置需要发报警邮件的状态,我只设置critical,减少邮件量
Types为states的类型,很多
Perio报警的时间段
如果你想延迟第一次报警的时间,可如下:
apply Notification "mail" to Service { import "generic-notification" command = "mail-notification" users = [ "icingaadmin" ] times.begin = 15m // delay first notification assign where service.name == "ping4" }
Tips:
When detecting a problem with a host/service Icinga re-checks theobject a number of times (based on the max_check_attempts and retry_intervalsettings) before sending notifications. This ensures that no unnecessarynotifications are sent for transient failures. During this time the object isin a SOFT state.After all re-checks have been executed and the object is still ina non-OK state the host/service switches to a HARD state and notifications are sent.
users.conf 用来定义报警和定义主机
object User "icingaadmin"{ import "generic-user" display_name = "Icinga 2 Admin" groups = [ "icingaadmins" ] email = "[email protected]" } object Host "xx" { display_name = "xx" address = "xx" groups = [ "cs" ] check_command = "hostalive" }
Object User 或Object Host是固定格式,后面的内容为自定义。
Host说明:
Import导入templates.conf里的模板
display_name 自定义
groups自定义,如果多个,用逗号隔开(但是是否每个都能用,有待确认)
address 可以是域名或者ip
check_command 检测主机的命令,这里用的hostalive,就是ping检测…
services.conf 定义服务(也可以给特别的服务单独写个xxx.conf)
objectService "ssh" { import "generic-service" check_command = "ssh" host_name= "hk" vars.ssh_port = "22221" }
针对单个主机的服务,可以用object Service的方式定义。
var.ssh_port这里是自定义参数的使用方式。var.为固定格式,后面跟参数名,参数名是在command-plugins.conf中定义的,等号后面是自定义的端口。
针对一个服务很多主机的情况,用如下apply service的方式定义:
applyService "total_procs" { import "generic-service" check_command = "nrpe" # use nrpe command to check vars.nrpe_command ="check_total_procs" #commandon client server assignwhere "es" in host.groups ignorewherehost.address == "" } apply Service "http 80" { import "generic-service" check_command = "http" # commandon monitor server which has argument “-H” assign where "vu" in host.groups ignore where host.address == "" }
用apply的方式,一定有关键词assign和ignore,后者可以为空,可以多行ignore(写在一行没成功)。
这里两个service定义,原理是一样的,都用插件,check_nrpe或者check_http,这里写的命令http或者nrpe已经定义在command-plugins.conf。
groups.confg 定义服务组或者主机组
object ServiceGroup "load" { display_name = "Load Checks" assignwhereservice.vars.nrpe_command== "check_load" } object ServiceGroup "ssh" { display_name = "Ssh Checks" assign where service.check_command== "ssh" } object HostGroup "es" { display_name = "es server" }
notifications.conf 应用报警(之前做了模板,现在是应用)
apply Notification "mail-icingaadmin" to Host { import "mail-host-notification" user_groups = [ "icingaadmins" ] assign where host.vars.sla == "24x7" }
apply Notification "mail-icingaadmin-5" to Service { import "5mins-notification" user_groups = [ "icingaadmins" ] assign where service.name == "ssh" assign where service.name == "check_system_5" assign where service.name == "zombie_procs" assign where service.name == "http 80" assign where service.name == "ssh" }
icinga2的配置的语法, 远比我上面写的复杂, 支持正则, 各种宏变量, 非常的灵活.
验证并加载配置
/etc/init.d/icinga2 reload (会自动检查配置)