ZABBIX使用LLD自动发现规则发现docker容器，并监测其运行状态（下）

利用ZABBIX自动发现监控功能，编写自定义功能脚本，使用外部检查的方式，无需安装zabbix agent客户端，实现自动获取服务器上运行的docker服务并监控其运行状态。

前提条件

已经部署好的zabbix监控系统

Zabbix服务器需安装sshpass命令

1. 编写脚本

1.1. 编写docker容器名获取脚本

vim docker_find.sh

#!/bin/bash
name_json="{\"data\":["
A='1'
passwd="$1"
monitor_ip="$2"
docker_name=$(sshpass -p "${passwd}" ssh -o StrictHostKeyChecking=no root@${monitor_ip} "docker ps -a" | grep -v 'STATUS' | awk '{printf $NF "\n"}')
docker_num=$(sshpass -p "${passwd}" ssh -o StrictHostKeyChecking=no root@${monitor_ip} "docker ps -a" | grep -v 'STATUS' | awk '{printf $NF "\n"}' | wc -l)
for do_name in ${docker_name}
do
  if [[ $A = ${docker_num} ]];then
    name_json+="{\"{#DOCKER_NAME}\": \"${do_name}\"}"
  else
    name_json+="{\"{#DOCKER_NAME}\": \"${do_name}\"},"
  fi
  ((A++))
done
name_json+="]}"
echo $name_json

这个脚本旨在输出一个json格式的数据，标准格式为：

{"data":[{"#aaa": "AAAA"},{....},{"#ccc": "CCC"}]}

ZABBIX会根据收到的这个数据，来创建监控项。

1.2. 编写docker容器运行状态监测脚本

vim docker_health_status.sh

#!/bin/bash
docker_name="$1"
passwd="$2"
monitor_ip="$3"
docker_num=$(sshpass -p "${passwd}" ssh -o StrictHostKeyChecking=no root@${monitor_ip} "docker ps -a" | grep -w "${docker_name}" | wc -l)
if [[ ! "${docker_num}" = 1 ]];then
  docker ps -a | grep -w "${docker_name}" > /tmp/.docker.txt
  cat /tmp/.docker.txt | while read docker_file
  do
    docker_if=$(echo ${docker_file}| awk '{printf $NF "\n"}')
    if [[ ${docker_if} = ${docker_name} ]];then
      docker_status=$(echo ${docker_file} | awk -F 'ago' '{printf $2 "\n"}' | awk '{printf $1 "\n"}')
      if [[ ${docker_status} = "Up" ]];then
        docker_health=$(echo ${docker_file} | grep "health")
        if [[ ${docker_health:-"nohealthcheck"} = "nohealthcheck" ]];then
          echo '1'
        else
          health_status=$(echo ${docker_file} | awk -F 'ago' '{printf $2 "\n"}' | awk -F ')' '{printf $NR "\n"}' | awk -F '(' '{printf $2 "\n"}')
          if [[ ${health_status} = 'healthy' ]];then
            echo '1'
          else
            echo '0'
          fi
        fi
      else
        echo '0'
      fi
    else
      continue
    fi
  done
else
  docker_status=$(sshpass -p "${passwd}" ssh -o StrictHostKeyChecking=no root@${monitor_ip} "docker ps -a" | grep -w "${docker_name}" | awk -F 'ago' '{printf $2 "\n"}' | awk '{printf $1 "\n"}')
  if [[ ${docker_status} = "Up" ]];then
    docker_health=$(sshpass -p "${passwd}" ssh -o StrictHostKeyChecking=no root@${monitor_ip} "docker ps -a" | grep -w "${docker_name}" | grep "health")
    if [[ ${docker_health:-"nohealthcheck"} = "nohealthcheck" ]];then
      echo '1'
    else
      health_status=$(sshpass -p "${passwd}" ssh -o StrictHostKeyChecking=no root@${monitor_ip} "docker ps -a" | grep -w "${docker_name}" | awk -F 'ago' '{printf $2 "\n"}' | awk -F ')' '{printf $NR "\n"}' | awk -F '(' '{printf $2 "\n"}')
      if [[ ${health_status} = 'healthy' ]];then
        echo '1'
      else
        echo '0'
      fi
    fi
  else
    echo '0'
  fi
fi

这个脚本是将上一个脚本获取的docker容器名，查找对应的容器，判断容器的运行状态（包括配置了docker健康检查探针）并返回一个值：0和1。

其中1表示运行正常，0表示异常