nagios系列(六)之nagios实现对服务器cpu温度的监控
1、安装硬件传感器监控软件sensors
yum install -y lm_sensors*
2、运行sensors-detect进行传感器检测
##一路回车即可
Do you want to overwrite /etc/sysconfig/lm_sensors? (YES/no):
Starting lm_sensors: loading module coretemp [ OK ]
Unloading i2c-dev... OK
3、运行sensors看是否能读取数据，如下像下面这样表示正常
# sensors
coretemp-isa-0000
Adapter: ISA adapter
ERROR: Can't get value of subfeature temp1_input: Can't read
Physical id 0: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp2_input: Can't read
Core 0: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp3_input: Can't read
Core 1: +0.0°C (high = +100.0°C, crit = +100.0°C)
coretemp-isa-0002
Adapter: ISA adapter
ERROR: Can't get value of subfeature temp1_input: Can't read
Physical id 1: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp2_input: Can't read
Core 0: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp3_input: Can't read
Core 1: +0.0°C (high = +100.0°C, crit = +100.0°C)
4、添加监控脚本vim /usr/local/nagios/libexec/check_cputemp
#!/bin/sh
#########check_cputemp###########
#date : May 2013
#Licence GPLv2
#by Barlow
#/usr/local/nagios/libexec/check_cputemp
#you can use NRPE to define service in nagios
#check_nrpe!check_cputemp
# Plugin return statements
STATE_OK=0
STATE_WARNING=1
STATE_CRITICAL=2
STATE_UNKNOWN=3
print_help_msg(){
$Echo "Usage: $0 -h to get help."
}
print_full_help_msg(){
$Echo "Usage:"
$Echo "$0 [ -v ] -m sensors -w cpuT -c cpuT"
$Echo "Sepicify the method to use the temperature data sensors."
$Echo "And the corresponding Critical value must greater than Warning value."
$Echo "Example:"
$Echo "${0} -m sensors -w 40 -c 50"
}
print_err_msg(){
$Echo "Error."
print_full_help_msg
}
to_debug(){
if [ "$Debug" = "true" ]; then
$Echo "$*" >> /var/log/check_sys_temperature.log.$$ 2>&1
fi
}
unset LANG
Echo="echo -e"
if [ $# -lt 1 ]; then
print_help_msg
exit 3
else
while getopts :vhm:w:c: OPTION
do
case $OPTION
in
v)
#$Echo "Verbose mode."
Debug=true
;;
m)
method=$OPTARG
;;
w)
WARNING=$OPTARG
;;
c)
CRITICAL=$OPTARG ;;
h)
print_full_help_msg
exit 3
;;
?)
$Echo "Error: Illegal Option."
print_help_msg
exit 3
;;
esac
done
if [ "$method" = "sensors" ]; then
use_sensors="true"
to_debug use_sensors
else
$Echo "Error. Must to sepcify the method to use sensors."
print_full_help_msg
exit 3
fi
to_debug All Values are \" Warning: "$WARNING" and Critical: "$CRITICAL" \".
fi
#########lm_sensors##################
if [ "$use_sensors" = "true" ]; then
sensorsCheckOut=`which sensors 2>&1`
if [ $? -ne 0 ];then
echo $sensorsCheckOut
echo Maybe you need to check your sensors.
exit 3
fi
to_debug Use $sensorsCheckOut to check system temperature
TEMP1=`sensors | head -3 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
TEMP2=`sensors | head -4 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
TEMP3=`sensors | head -5 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
TEMP4=`sensors | head -6 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
##温度的取数根据你cpu的核数确定，我的是四核，所以取TEMP1-4个CPU温度数并计算平均值
SUM=$(( $TEMP1 + $TEMP2 + $TEMP3 + $TEMP4 ))
TEMP=$(($SUM/4))
if [ -z "$TEMP" ] ; then
$Echo "No Data been get here. Please confirm your ARGS and re-check it with Verbose mode, then to check the log."
exit 3
fi
to_debug temperature data is $TEMP
else
$Echo "Error. Must to sepcify the method to use sensors"
print_full_help_msg
exit 3
fi
######### Comparaison with the warnings and criticals thresholds given by user############
CPU_TEMP=$TEMP
#if [ "$WARNING" != "0" ] || [ "$CRITICAL" != "0" ]; then
if [ "$CPU_TEMP" -gt "$CRITICAL" ] && [ "$CRITICAL" != "0" ]; then
STATE="$STATE_CRITICAL"
STATE_MESSAGE="CRITICAL"
to_debug $STATE , Message is $STATE_MESSAGE
elif [ "$CPU_TEMP" -gt "$WARNING" ] && [ "$WARNING" != "0" ]; then
STATE="$STATE_WARNING"
STATE_MESSAGE="WARNING"
to_debug $STATE , Message is $STATE_MESSAGE
else
STATE="$STATE_OK"
STATE_MESSAGE="OK"
to_debug $STATE , Message is $STATE_MESSAGE
fi
##返回值中注意要包含性能数据，即采用|分隔的后半部数据，且数据单位不能包含中文，否则使用PNP等绘图软件无法正常绘图。
echo "The TEMPERATURE "$STATE_MESSAGE" "-" The CPU's Temperature is "$CPU_TEMP" ℃ ! | 温度=`echo $CPU_TEMP`Celsius;$WARNING;$CRITICAL"
exit $STATE
5、赋予脚本执行权限：
chmod +x /usr/local/nagios/libexec/check_cputemp
6、配置vim /usr/local/nagios/etc/nrpe.cfg，添加如下一行：
echo "command[check_cputemp]=/usr/local/nagios/libexec/check_cputemp -m sensors -w 38 -c 45" >>/usr/local/nagios/etc/nrpe.cfg
重新启动客户端nrpe服务
-w 表示警告值，-c表示关键（紧急）值，自行根据实际情况调整
注意：以上六步均在被监控机上完成。
在客户端测试是否ok，虚拟机测试不成功，需要在物理机上实现
# /usr/local/nagios/libexec/check_cputemp -m sensors -w 38 -c 45
The TEMPERATURE OK - The CPU's Temperature is 14 ℃ ! | 温度=14Celsius;38;45
服务端执行测试：
/usr/local/nagios/libexec/check_nrpe -H 192.168.8.93 -c check_cputemp
7、在Nagios服务端配置服务：
define service{
use generic-service
host_name 需要被监控的hostname
service_description CPU Temperature
check_command check_nrpe!check_cputemp
}
yum install -y lm_sensors*
2、运行sensors-detect进行传感器检测
##一路回车即可
Do you want to overwrite /etc/sysconfig/lm_sensors? (YES/no):
Starting lm_sensors: loading module coretemp [ OK ]
Unloading i2c-dev... OK
3、运行sensors看是否能读取数据，如下像下面这样表示正常
# sensors
coretemp-isa-0000
Adapter: ISA adapter
ERROR: Can't get value of subfeature temp1_input: Can't read
Physical id 0: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp2_input: Can't read
Core 0: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp3_input: Can't read
Core 1: +0.0°C (high = +100.0°C, crit = +100.0°C)
coretemp-isa-0002
Adapter: ISA adapter
ERROR: Can't get value of subfeature temp1_input: Can't read
Physical id 1: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp2_input: Can't read
Core 0: +0.0°C (high = +100.0°C, crit = +100.0°C)
ERROR: Can't get value of subfeature temp3_input: Can't read
Core 1: +0.0°C (high = +100.0°C, crit = +100.0°C)
4、添加监控脚本vim /usr/local/nagios/libexec/check_cputemp
#!/bin/sh
#########check_cputemp###########
#date : May 2013
#Licence GPLv2
#by Barlow
#/usr/local/nagios/libexec/check_cputemp
#you can use NRPE to define service in nagios
#check_nrpe!check_cputemp
# Plugin return statements
STATE_OK=0
STATE_WARNING=1
STATE_CRITICAL=2
STATE_UNKNOWN=3
print_help_msg(){
$Echo "Usage: $0 -h to get help."
}
print_full_help_msg(){
$Echo "Usage:"
$Echo "$0 [ -v ] -m sensors -w cpuT -c cpuT"
$Echo "Sepicify the method to use the temperature data sensors."
$Echo "And the corresponding Critical value must greater than Warning value."
$Echo "Example:"
$Echo "${0} -m sensors -w 40 -c 50"
}
print_err_msg(){
$Echo "Error."
print_full_help_msg
}
to_debug(){
if [ "$Debug" = "true" ]; then
$Echo "$*" >> /var/log/check_sys_temperature.log.$$ 2>&1
fi
}
unset LANG
Echo="echo -e"
if [ $# -lt 1 ]; then
print_help_msg
exit 3
else
while getopts :vhm:w:c: OPTION
do
case $OPTION
in
v)
#$Echo "Verbose mode."
Debug=true
;;
m)
method=$OPTARG
;;
w)
WARNING=$OPTARG
;;
c)
CRITICAL=$OPTARG ;;
h)
print_full_help_msg
exit 3
;;
?)
$Echo "Error: Illegal Option."
print_help_msg
exit 3
;;
esac
done
if [ "$method" = "sensors" ]; then
use_sensors="true"
to_debug use_sensors
else
$Echo "Error. Must to sepcify the method to use sensors."
print_full_help_msg
exit 3
fi
to_debug All Values are \" Warning: "$WARNING" and Critical: "$CRITICAL" \".
fi
#########lm_sensors##################
if [ "$use_sensors" = "true" ]; then
sensorsCheckOut=`which sensors 2>&1`
if [ $? -ne 0 ];then
echo $sensorsCheckOut
echo Maybe you need to check your sensors.
exit 3
fi
to_debug Use $sensorsCheckOut to check system temperature
TEMP1=`sensors | head -3 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
TEMP2=`sensors | head -4 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
TEMP3=`sensors | head -5 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
TEMP4=`sensors | head -6 | tail -1 | gawk '{print $3}' | grep -o [0-9][0-9]`
##温度的取数根据你cpu的核数确定，我的是四核，所以取TEMP1-4个CPU温度数并计算平均值
SUM=$(( $TEMP1 + $TEMP2 + $TEMP3 + $TEMP4 ))
TEMP=$(($SUM/4))
if [ -z "$TEMP" ] ; then
$Echo "No Data been get here. Please confirm your ARGS and re-check it with Verbose mode, then to check the log."
exit 3
fi
to_debug temperature data is $TEMP
else
$Echo "Error. Must to sepcify the method to use sensors"
print_full_help_msg
exit 3
fi
######### Comparaison with the warnings and criticals thresholds given by user############
CPU_TEMP=$TEMP
#if [ "$WARNING" != "0" ] || [ "$CRITICAL" != "0" ]; then
if [ "$CPU_TEMP" -gt "$CRITICAL" ] && [ "$CRITICAL" != "0" ]; then
STATE="$STATE_CRITICAL"
STATE_MESSAGE="CRITICAL"
to_debug $STATE , Message is $STATE_MESSAGE
elif [ "$CPU_TEMP" -gt "$WARNING" ] && [ "$WARNING" != "0" ]; then
STATE="$STATE_WARNING"
STATE_MESSAGE="WARNING"
to_debug $STATE , Message is $STATE_MESSAGE
else
STATE="$STATE_OK"
STATE_MESSAGE="OK"
to_debug $STATE , Message is $STATE_MESSAGE
fi
##返回值中注意要包含性能数据，即采用|分隔的后半部数据，且数据单位不能包含中文，否则使用PNP等绘图软件无法正常绘图。
echo "The TEMPERATURE "$STATE_MESSAGE" "-" The CPU's Temperature is "$CPU_TEMP" ℃ ! | 温度=`echo $CPU_TEMP`Celsius;$WARNING;$CRITICAL"
exit $STATE
5、赋予脚本执行权限：
chmod +x /usr/local/nagios/libexec/check_cputemp
6、配置vim /usr/local/nagios/etc/nrpe.cfg，添加如下一行：
echo "command[check_cputemp]=/usr/local/nagios/libexec/check_cputemp -m sensors -w 38 -c 45" >>/usr/local/nagios/etc/nrpe.cfg
重新启动客户端nrpe服务
-w 表示警告值，-c表示关键（紧急）值，自行根据实际情况调整
注意：以上六步均在被监控机上完成。
在客户端测试是否ok，虚拟机测试不成功，需要在物理机上实现
# /usr/local/nagios/libexec/check_cputemp -m sensors -w 38 -c 45
The TEMPERATURE OK - The CPU's Temperature is 14 ℃ ! | 温度=14Celsius;38;45
服务端执行测试：
/usr/local/nagios/libexec/check_nrpe -H 192.168.8.93 -c check_cputemp
7、在Nagios服务端配置服务：
define service{
use generic-service
host_name 需要被监控的hostname
service_description CPU Temperature
check_command check_nrpe!check_cputemp
}
保存后重启nagios服务
