服务器批量健康检测(ansible+python+shell)


一、shell脚本、统计cpu、内存、磁盘、I/O、连接数、历史登录用户、计划任务等信息(health_check.sh)

#!/bin/bash
#Result_file=/usr/local/script/Health_check/test/result.txt
Result_file=/tmp/system_tmp/systeminfo.txt   #检测结果存储文件
[ -f $Result_file ]|| mkdir /tmp/system_tmp/
>$Result_file



disk_judge=75      #磁盘使用率,磁盘使用率大于75%的检测会失败
cpu_judge=`cat /proc/cpuinfo| grep "processor"| wc -l|awk '{print $1*"0.50"}'` #cpu负载,cpu负载大于"cpu核数*50%"会检测失
Mem_judge=0.80    #内存使用率大于80%会检测失败
Io_judge=80      #I/O请求占用的CPU百分比,值越高,说明I/O越慢,大于80%则不通过,该值仅做参考
Last_judge="^172|^10"   #ssh登录情况,排除内网登录
Cron_judge="Scp_Tool|first_pvuv2|logservicerestart|cut_log|log_cut|log_gzip|clear_log" #计划任务查看,排除日志切割脚本




Interval(){

if [ $1 = a ];then

echo "********************************$2********************************" >>$Result_file

elif [ $1 = b ];then
echo "----------------------------------------" >>$Result_file
else
echo "----------------------------------------" >>$Result_file
fi

}


Disk_Check(){
Result="PASS"
Disk_use=`df -hP|grep -Ev 'Filesystem|tmpfs|boot'|awk '{print $5}'|xargs | sed 's/ /,/g'|sed 's/%//g'`
array=(${Disk_use//\,/ })
for i in "${!array[@]}"; do
    if [ ${array[i]} -gt $disk_judge ];then
      Result=FAILED
    fi
done

Interval a "DISK INFO"
df -h >>$Result_file
Interval b
echo "Check_Result:$Result"  >>$Result_file
Interval b
}

Cpu_Check(){
Result="PASS"
Load_average=`uptime|awk -F "average:" '{print$2}'`
array=(${Load_average//\,/ })
for i in "${!array[@]}"; do
    compare=`awk -v num1=${array[i]} -v num2=$cpu_judge 'BEGIN{print(num1>num2)?"0":"1"}'`
    if [ $compare -eq 0 ];then
      Result=FAILED
    fi
done



Interval a "CPU INFO"
iostat |grep -A1 avg-cpu >>$Result_file
echo "load_average:$Load_average">>$Result_file
Interval b
echo "Check_Result:$Result"  >>$Result_file
Interval b

}

Connection_Check(){

Interval a "CONNECTTION INFO"
netstat -n | awk '/^tcp/ {++S[$NF]} END {for(a in S) print a, S[a]}' >>$Result_file
Interval b
echo "Check_Result:$Result"  >>$Result_file
Interval b

}

Mem_Check(){
Result="PASS"
Mem_use=`free -m|grep Mem|awk '{print "1"-($4+$6+$7)/$2}'`
compare=`awk -v num1=$Mem_use -v num2=$Mem_judge 'BEGIN{print(num1>num2)?"0":"1"}'`
if [ $compare -eq 0 ];then
  Result=FAILED
fi

Interval a "Memory INFO"
free -m >>$Result_file
Interval b
echo "Check_Result:$Result"  >>$Result_file
Interval b

}
Io_Check(){


Result="PASS"
Interval a "IO INFO"
sar -dp 1 3|grep  "Average" >>$Result_file
Io_use=`cat /tmp/system_tmp/systeminfo.txt|grep "Average:"|grep -v util |awk  '{print $10}'|xargs | sed 's/ /,/g'`

array=(${Io_use//\,/ })
for i in "${!array[@]}"; do
    compare1=`awk -v num1=${array[i]} -v num2=$Io_judge 'BEGIN{print(num1>num2)?"0":"1"}'`
    if [ $compare -eq 0 ];then
      Result=FAILED
    fi
done

Interval b
echo "Check_Result:$Result"  >>$Result_file
Interval b
}


Last_Check(){
Result="PASS"
lastcheck=`last|grep -Ev "tty|reboot|wtmp" |awk '{print$3}'|grep -Ev $Last_judge|sed '/^$/d'|wc -l`
if [ $lastcheck -ne 0 ];then
  Result=FAILED
fi 

Interval a "LAST INFO"
last -an 10 | uniq -f 9 >>$Result_file
Interval b
echo "Check_Result:$Result"  >>$Result_file
Interval b
}


Crontab_Check(){
Result="PASS"
crocheck1=`cat /etc/crontab |grep -Ev "^#|HOME|MAILTO|PATH|SHELL|$Cron_judge"|sed '/^$/d'|wc -l`
crocheck2=`cat /etc/passwd | cut -f 1 -d : |xargs -I {} crontab -l -u {}|grep -Ev $Cron_judge|wc -l`
crocheck3=`cat /etc/passwd | cut -f 1 -d : |xargs -I {} crontab -l -u {}|wc -l`
if [ $crocheck1 -ne 0 ]||[ $crocheck2 -ne 0 ];then
  Result=FAILED
fi

Interval a "CRONTAB INFO"
if [ $crocheck3 -eq 0 ]&&[ $crocheck1 -eq 0 ];then
echo "no crontab for all" >>$Result_file

else
cat /etc/passwd | cut -f 1 -d : |xargs -I {} crontab -l -u {} >>$Result_file
cat /etc/crontab |grep -Ev "^#|HOME|MAILTO|PATH|SHELL|$Cron_judge"|sed '/^$/d' >>$Result_file 
fi
Interval b
echo "Check_Result:$Result"  >>$Result_file
Interval b
}


Disk_Check
Cpu_Check
Connection_Check
Mem_Check
Io_Check
Last_Check
Crontab_Check


二、使用ansible批量执行并返回执行结果(systemcheck.yml)

- name: create dir
  file:
    path: /tmp/system_tmp
    state: directory
    mode: 0755
- name: copy cpu_mem.sh to remote machine
  copy:
    src: "health_check.sh"
    dest: /tmp/system_tmp/health_check.sh
    owner: root
    group: root
    mode: a+x
    remote_src: False
- name: write cpu,Mem to /tmp/systeminfo.txt
  shell: "sh /tmp/system_tmp/health_check.sh"
  
- name: store file into /tmp/checklog/
  fetch:
    src: /tmp/system_tmp/systeminfo.txt
    dest: /tmp/checklog/
    mode: 0644


三、每台服务器返回的结果进行分析 并格式化输出

#!/usr/bin/python3.6
# -*- coding: utf-8 -*-

import os
import datetime
import docx
from docx import Document
from docx.enum.text import WD_PARAGRAPH_ALIGNMENT
from docx.shared import Pt
from docx.shared import RGBColor


now_time = datetime.datetime.now()  # 当前时间

Result_file = "/usr/local/script/Health_check/hostcheck_result.docx"  #检查结果存储文件
#读取ansible返回的主机信息,以文件存储
Path="/tmp/checklog/"
Listdir=os.listdir(Path)

#统计检查主机总数
Total_count = 0
#统计检查不通过主机数
Fail_count = 0
#统计检查不通过主机IP存入Fail_list列表
Fail_list=[]
for Dir in Listdir:             #循环目录,获取每个ansible返回的每个文件
  File="/tmp/checklog/"+Dir+"/tmp/system_tmp/systeminfo.txt"
  with open(File,"r",encoding="utf-8") as f:
    for line in f:
      if "Check_Result:FAILED" in line:
        Fail_list.append(Dir)
        Fail_count = Fail_count + 1        
  Total_count = Total_count + 1


Title="系统健康检测报告"
Check_user="封孝立"
Check_time=now_time.strftime("%Y-%m-%d")
Check_hosts=int(Total_count)
Pass_host=int(Total_count)-int(Fail_count)
Failed_host=int(Fail_count)



# 声明一个word对象
doc = Document()

#title
Report_Title = doc.add_heading(Title, 0)

# 设置Title水平居中
paragraph_format = Report_Title.paragraph_format
paragraph_format.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER


Checkinfo = '''
检测人:%s

检测时间:%s

检测项:cpu、内存、磁盘、I/O、连接数、历史登录用户、计划任务

检测标准:cpu负载>Cpu_core*0.5,内存使用率>0.8 ,磁盘使用率>0.75, IO>util*0.8,ssh外网登录

检测主机总数:%d

检测通过主机数:%d

检测故障主机数:%d

故障主机IP列表:
'''%(Check_user,Check_time,Check_hosts,Pass_host,Failed_host)

#输入检测基本信息
Basic_info = doc.add_paragraph(Checkinfo)

Fail_list = list(set(Fail_list))  #检查不通过去重IP
for failip in Fail_list:
  Failip = doc.add_paragraph('').add_run(u''+failip.strip('\n'))
  Failip.font.color.rgb = RGBColor(250,0,0)

#分页
doc.add_page_break()




for Dir in Listdir:             #循环目录,获取每个ansible返回的每个文件
  #将IP设置成标题,并修改字号
  Ip_title=doc.add_heading('',0).add_run(u"HOST IP:"+Dir)
  Ip_title.font.size = Pt(20)

  File="/tmp/checklog/"+Dir+"/tmp/system_tmp/systeminfo.txt"
  with open(File,"r",encoding="utf-8") as f:
    for line in f:
      if "Check_Result:PASS" in line:
        paragraph = doc.add_paragraph('').add_run(u''+line.strip('\n'))
        paragraph.font.size = Pt(12)
        paragraph.font.color.rgb = RGBColor(0,255,0)
#        paragraph.bold = True
      elif "Check_Result:FAILED" in line:
        paragraph = doc.add_paragraph('').add_run(u''+line.strip('\n'))
        paragraph.font.size = Pt(12)
        paragraph.font.color.rgb = RGBColor(250,0,0)
#        paragraph.bold = True
      else:
        paragraph = doc.add_paragraph(line.strip('\n'))
  doc.add_page_break()  


doc.save(Result_file)


四、检测结果示例

1602315137(1).jpg

1602315460(1).jpg

1602315536(1).jpg

1602315549(1).jpg










猜你喜欢

转载自blog.51cto.com/fengxiaoli/2541011