在本教程中，我将展示如何编写shell脚本以执行Linux服务器运行状况检查。
此脚本收集主机名，内核版本，正常运行时间，CPU，内存和磁盘使用等系统信息和状态。

该脚本使用主机名，正常运行时间，WHO，MPSTAT，LSCPU，PS，Top，DF，免费，BC命令来获取系统信息和切割，Grep，AWK和SED进行文本处理。
脚本的输出是将在当前目录中生成的文本文件。
设置变量以提供脚本可以发送报告文件的电子邮件地址。
除系统状态外，脚本将检查CPU负载和文件系统大小的预定义阈值。

请记住：确保我们拥有上述所有命令，正确输出所有结果。

要监控服务器健康的脚本

让我们仔细检查有关此脚本的详细信息，有助于监视Linux服务器。

#tell which shell to use
#!/bin/bash
#Here we put email address to send email with report. If no email provided – log file will be just saved.
EMAIL='[email protected]'
#We will create function to easily manage what to do with output.
function sysstat {
#Print header, hostname (hostname command used), Kernel version (uname -r) , Uptime (from uptime command) and Last reboot time (from who command)
echo -e "
#!/bin/bash
function sysstat {
echo -e "Health Check Report (CPU,Process,Disk Usage, Memory)#hostname command returns hostname
Hostname : `hostname`
#uname command with key -r returns Kernel version
Kernel Version : `uname -r`
#uptime command used to get uptime, and with sed command we cat process output to get only uptime.
Uptime : `uptime | sed 's/.*up \([^,]*\), .*//'`
#who command is used to get last reboot time, awk for processing output
Last Reboot Time : `who -b | awk '{print ,}'`
*
CPU Load - > Threshold < 1 Normal > 1 Caution , > 2 Unhealthy
*
"
#here we check if mpstat command is in our system
MPSTAT=`which mpstat`
#here we get exit code from previous command
MPSTAT=$?
#if exit status in not 0, this means that mpstat command is not found (or not exist in our system)
if [ $MPSTAT != 0 ]
then
echo "Please install mpstat!"
echo "On Debian based systems:"
echo "sudo apt-get install sysstat"
echo "On RHEL based systems:"
echo "yum install sysstat"
else
echo -e ""
#here we check in same way if lscpu installed
LSCPU=`which lscpu`
LSCPU=$?
if [ $LSCPU != 0 ]
then
RESULT=$RESULT" lscpu required to procedure accurate results"
else
#if we have lscpu installed, we can get number of CPU's on our system and get statistic for each using mpstat command.
cpus=`lscpu | grep -e "^CPU(s):" | cut -f2 -d: | awk '{print }'`
i=0
#here we make loop to get and print CPU usage statistic for each CPU.
while [ $i -lt $cpus ]
do
#here we get statistic for CPU and print it. Awk command help to do this, since output doesn't allow this to do with grep. AWK check if third value is equal to variable $i (it changes from 0 to number of CPU), and print %usr value for this CPU
echo "CPU$i : `mpstat -P ALL | awk -v var=$i '{ if ( == var ) print  }' `"
#here we increment $i variable for loop
let i=$i+1
done
fi
echo -e "
#here with uptime command we get load average for system, and cut command helps to process result.
Load Average : `uptime | awk -F'load average:' '{ print  }' | cut -f1 -d,`
#same as before, but with awk command we check if system is Normal (if value less than 1, Caution (if between 1 and 2) and Unhealthy.
Heath Status : `uptime | awk -F'load average:' '{ print  }' | cut -f1 -d, | awk '{if ( > 2) print
"Unhealthy"; else if ( > 1) print "Caution"; else print "Normal"}'`
"
fi
echo -e "
**
Process
**
Top memory using processs/application
PID %MEM RSS COMMAND
#with ps command we get list of processes,  awk show only needed columns. After with sort command we sort it by third column and we need only top 10, that why we used head command
`ps aux | awk '{print , , , }' | sort -k3rn | head -n 10`
Top CPU using process/application
#with top command we get top CPU using processes, and with combination of head and tail we get top 10.
`top b -n1 | head -17 | tail -11`
**
Disk Usage - > Threshold < 90 Normal > 90% Caution > 95 Unhealthy
**
"
#we get disk usage with df command. -P key used to have postfix like output (there was problems with network shares, etc and -P resolve this problems). We print output to temp file to work with info more than one.
df -Pkh | grep -v 'Filesystem' > /tmp/df.status
#We create loop to process line by line from df.status
while read DISK
do
#here we get line from df.status and print result formatted with awk command
LINE=`echo $DISK | awk '{print ,"\t",,"\t",," used","\t",," free space"}'`
echo -e $LINE
echo
done < /tmp/df.status
echo -e "
Heath Status"
echo
#here almost same loop, but we check disk usage, and print Normal if value less 90, Caution if between 90 and 95, and Unhealthy if greater than 95)
while read DISK
do
USAGE=`echo $DISK | awk '{print }' | cut -f1 -d%`
if [ $USAGE -ge 95 ]
then
STATUS='Unhealthy'
elif [ $USAGE -ge 90 ]
then
STATUS='Caution'
else
STATUS='Normal'
fi
LINE=`echo $DISK | awk '{print ,"\t",}'`
#here we print result with status
echo -ne $LINE "\t\t" $STATUS
echo
done < /tmp/df.status
#here we remove df.status file
rm /tmp/df.status
#here we get Total Memory, Used Memory, Free Memory, Used Swap and Free Swap values and save them to variables.
TOTALMEM=`free -m | head -2 | tail -1| awk '{print }'`
#All variables like this is used to store values as float (we are using bc to make all mathematics operations, since without bc all values will be integer). Also we use if to add zero before value, if value less than 1024, and result of dividing will be less than 1.
TOTALBC=`echo "scale=2;if($TOTALMEM 0) print 0;$TOTALMEM/1024"| bc -l`
USEDMEM=`free -m | head -2 | tail -1| awk '{print }'`
USEDBC=`echo "scale=2;if($USEDMEM 0) print 0;$USEDMEM/1024"|bc -l`
FREEMEM=`free -m | head -2 | tail -1| awk '{print }'`
FREEBC=`echo "scale=2;if($FREEMEM 0) print 0;$FREEMEM/1024"|bc -l`
TOTALSWAP=`free -m | tail -1| awk '{print }'`
TOTALSBC=`echo "scale=2;if($TOTALSWAP 0) print 0;$TOTALSWAP/1024"| bc -l`
USEDSWAP=`free -m | tail -1| awk '{print }'`
USEDSBC=`echo "scale=2;if($USEDSWAP 0) print 0;$USEDSWAP/1024"|bc -l`
FREESWAP=`free -m | tail -1| awk '{print }'`
FREESBC=`echo "scale=2;if($FREESWAP 0) print 0;$FREESWAP/1024"|bc -l`
echo -e "

Memory

Physical Memory
Total\tUsed\tFree\t%Free
# as we get values in GB, also we get % of usage dividing Free by Total
${TOTALBC}GB\t${USEDBC}GB \t${FREEBC}GB\t$(($FREEMEM * 100/$TOTALMEM ))%
Swap Memory
Total\tUsed\tFree\t%Free
#Same as above – values in GB, and in same way we get % of usage
${TOTALSBC}GB\t${USEDSBC}GB\t${FREESBC}GB\t$(($FREESWAP * 100/$TOTALSWAP ))%
"
}
#here we make filename value, using hostname, and date.
FILENAME="health-`hostname`-`date +%y%m%d`-`date +%H%M`.txt"
#here we run function and save result to generated filename
sysstat > $FILENAME
#here we print output to user.
echo -e "Reported file $FILENAME generated in current directory." $RESULT
#here we check if user provide his email address to send email
if [ "$EMAIL" != '' ]
then
#if email proviced – we check if we have mailx command to send email
STATUS=`which mail`
#if mailx command not exist on system (previous command returned non-zero exit code we warn user that mailx is not installed
if [ "$?" != 0 ]
then
echo "The program 'mail' is currently not installed."
#if mailx installed, we send email with report to user
else
cat $FILENAME | mail -s "$FILENAME" $EMAIL
fi
fi

服务器健康报告

[root@theitroad script]# cat health-theitroad.com-140831-0909.txt | moreHealth Check Report (CPU,Process,Disk Usage, Memory)Hostname : theitroad.com
Kernel Version : 3.15.4-x86_64-theitroad342
Uptime : 7 days
Last Reboot Time : 2014-08-27 08:46
*
CPU Load - > Threshold < 1 Normal > 1 Caution , > 2 Unhealthy
*
CPU0 : 0.06
Load Average : 0.00
Heath Status : Normal
*
Process
*
=> Top memory using processs/application
PID %MEM RSS COMMAND
1361 12.6 127896 /usr/lib/systemd/systemd-journald
1642 7.1 72252 /usr/sbin/rsyslogd
1644 1.9 20148 /usr/bin/python
2340 1.8 19092 /sbin/dhclient
1634 1.4 14748 /usr/sbin/NetworkManager
31410 0.8 8724 sshd:
31441 0.7 7888 sshd:
31432 0.7 7784 sshd:
2558 0.5 5988 /usr/sbin/sshd
1 0.5 5412 /sbin/init
=> Top CPU using process/application
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1 root 20 0 98572 5412 2796 S 0.0 0.5 0:11.82 systemd
2 root 20 0 0 0 0 S 0.0 0.0 0:00.00 kthreadd
3 root 20 0 0 0 0 S 0.0 0.0 0:00.51 ksoftirqd/0
5 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/0:0H
6 root 20 0 0 0 0 S 0.0 0.0 0:03.33 kworker/u2:0
7 root 20 0 0 0 0 S 0.0 0.0 0:04.17 rcu_sched
8 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcu_bh
9 root rt 0 0 0 0 S 0.0 0.0 0:00.00 migration/0
10 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 khelper
11 root 20 0 0 0 0 S 0.0 0.0 0:00.00 kdevtmpfs
*
Disk Usage - > Threshold < 90 Normal > 90% Caution > 95 Unhealthy
*
/dev/root/11% used 13G free space
devtmpfs /dev 0% used 495M free space
tmpfs /dev/shm 0% used 496M free space
tmpfs /run 1% used 495M free space
tmpfs /sys/fs/cgroup 0% used 496M free space
tmpfs /tmp 0% used 496M free space
Heath Status
/dev/root/Normal
devtmpfs /dev Normal
tmpfs /dev/shm Normal
tmpfs /run Normal
tmpfs /sys/fs/cgroup Normal
tmpfs /tmp Normal
*
Memory
*
=> Physical Memory
Total Used Free %Free
0.96GB 0.46GB 0.50GB 52%
=> Swap Memory
Total Used Free %Free
0.24GB 0GB 0.24GB 100%
[root@theitroad script]#