linux下对服务器性能监控shell脚本

  1 #!/bin/bash
  2 
  3 #提取本服务器的IP地址信息
  4 ENO1=`ifconfig | sed -n '1,1p' | awk -F ' ' '{print $1}'`
  5 IP=`ifconfig ${ENO1}| grep "inet addr" | cut -f 2 -d ":" | cut -f 1 -d " "`
  6 echo 'eno: '${ENO1}
  7 echo 'IP: '${IP}
  8 
  9 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
 10 # 各个监控警告值
 11 WARN_LOAD=1.1
 12 WARN_CPU=0.95
 13 WARN_MEM=0.93
 14 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
 15 
 16 
 17 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
 18 CPU=0.0
 19 LOAD1=0.0
 20 LOAD5=0.0
 21 LOAD15=0.0
 22 MEM=0.0
 23 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
 24 
 25 
 26 send_mail() {
 27     echo 'send mail '  "${1}" "${2}"
 28 }
 29 
 30 
 31 # 1、监控系统负载的变化情况.
 32 system_load_func()
 33 {
 34     cpu_num=`grep -c 'model name' /proc/cpuinfo`
 35 
 36     load=$(uptime | awk -F 'load average: ' '{print $2}')
 37     load_1=$(echo $load | awk -F ', ' '{print $1}')
 38     load_5=$(echo $load | awk -F ', ' '{print $2}')
 39     load_15=$(echo $load | awk -F ', ' '{print $3}')
 40 
 41     #计算当前系统单个核心平均负载值，结果小于1.0时前面个位数补0。
 42     LOAD1=`echo "scale=2;a=${load_1}/${cpu_num};if(length(a)==scale(a)) print 0;print a" | bc`
 43     LOAD5=`echo "scale=2;a=${load_5}/${cpu_num};if(length(a)==scale(a)) print 0;print a" | bc`
 44     LOAD15=`echo "scale=2;a=${load_15}/${cpu_num};if(length(a)==scale(a)) print 0;print a" | bc`
 45 
 46     if [ `echo "${LOAD15} > ${WARN_LOAD}" | bc` -eq 1 ]
 47     then
 48         send_mail "${IP}服务器15分钟的系统平均负载为${LOAD15}, 超过警戒值${WARN_LOAD}, 请立即处理!" "$IP 服务器系统负载告警!"
 49     fi
 50 }
 51 
 52 
 53 # 2、监控系统cpu的情况.
 54 cpu_func()
 55 {
 56     #取当前空闲cpu百份比值(只取整数部分)
 57     cpu_idle=`top -b -d 0.1 -n 2 | grep Cpu | tail -n 1 | awk '{print $8}' | cut -f 1 -d "."`
 58     CPU=0`echo "scale=2; (100 - ${cpu_idle}) / 100" | bc`
 59 
 60     if [ `echo "${CPU} > ${WARN_CPU}" | bc` -eq 1 ]
 61     then
 62         send_mail "${IP}服务器cpu使用${CPU},请及时处理." "$IP 服务器CPU告警"
 63     fi
 64 }
 65 
 66 
 67 #3. 监控物理内存
 68 mem_func()
 69 {
 70     mem=$(free -m | sed -n '2,2p')
 71     mem_total=`echo ${mem} | awk -F ' ' '{print $2}'`
 72     mem_used=`echo ${mem} | awk -F ' ' '{print $3}'`
 73     # mem_free=`echo ${mem} | awk -F ' ' '{print $4}'`
 74 
 75     if [ $mem_used -ne 0 ]
 76     then
 77         MEM=0`echo "scale=2; ${mem_used} / ${mem_total}" | bc`
 78         if [ `echo "${MEM} > ${WARN_MEM}" | bc` -eq 1 ]
 79         then
 80             send_mail "$IP服务器物理内存已使用 ${MEM},请及时处理." "$IP 服务器内存告警"
 81         fi
 82     fi
 83 }
 84 
 85 
 86 CNT=1
 87 print_info() {
 88     echo "# # # # # # # # # # # [${CNT}] # # # # # # # # # # # # # #"
 89     echo cpu: ${CPU}
 90     echo load1: ${LOAD1}
 91     echo load5: ${LOAD5}
 92     echo load15: ${LOAD15}
 93     echo mem: ${MEM}
 94     echo "# # # # # # # # # # # # # # # # # # # # # # # # # # #"
 95     (( CNT += 1 ))
 96     # exit 0
 97 }
 98 
 99 
100 while true
101 do
102     system_load_func
103     cpu_func
104     mem_func
105     print_info
106     sleep 5
107 done

注: 需要下载bc命令:

sudo apt-get install bc

这里我只在脚本里面写了获取各个监控信息的数据, 发送邮件和数据写入数据库这些我省略掉了. 发送邮件我使用的是python, 通过该shell脚本调用python发送邮件.(因为用linux自带的我发现会出现很多垃圾邮件, 怎么解决我没有去研究). 得到的数据我会写入数据库记录, 在这里为了通用性, 我把写入数据库部分省略掉了.