阿里云环境迁移记录 - 服务监控及报警

Posted 2021-02-10

服务监控的方案有很多，譬如naigos，zabbix这种，不但可以监控服务，还可以监控cpu、内存、磁盘、网络流量、服务端口等，关于naigos和zabbix的搭建配置，需要另外篇幅介绍，这里使用服务器自身的定时任务+脚本+邮件功能完成一个简单的监控。

Part1 邮件服务搭建

安装mailx

yum -y install mailx

############################
##qq个人邮箱配置
############################

vim /etc/mail.rc
添加如下配置：

set [email protected]
set smtp=smtps://smtp.qq.com:465
set [email protected]
set smtp-auth-password=你的 QQ 邮箱授权码 （登录qq邮箱到账户设置中，打开smtp服务时，提示的验证码，此验证码非密码。）
set smtp-auth=login
#set smtp-use-starttls 这里是不需要配置的，很多地方没说明，配置了反而会验证失败，所以我注释掉；
set ssl-verify=ignore
set nss-config-dir=/root/.certs

##创建证书

mkdir -p /root/.certs/
cd /root/.certs/
echo -n | openssl s_client -connect smtp.qq.com:465 | sed -ne ‘/-BEGIN CERTIFICATE-/,/-END CERTIFICATE-/p‘ > ~/.certs/qq.crt
certutil -A -n "GeoTrust SSL CA" -t "C,," -d ~/.certs -i ~/.certs/qq.crt
certutil -A -n "GeoTrust Global CA" -t "C,," -d ~/.certs -i ~/.certs/qq.crt
certutil -L -d /root/.certs
##认证
certutil -A -n "GeoTrust SSL CA - G3" -t "Pu,Pu,Pu" -d ./ -i qq.crt

#返回如下提示即可：
Notice: Trust flag u is set automatically if the private key is present.

#发送主题为“邮箱测试”，内容为当前目录下 message_fiel.txt 文件内容到 [email protected] 邮箱。
mailx -s "邮箱测试" [email protected] < message_file.txt

############################
##qq企业邮箱配置
############################

vim /etc/mail.rc
添加如下配置:

set [email protected]
set smtp=smtps://smtp.exmail.qq.com:465
set [email protected]
set smtp-auth-password=*****(登录密码，不同于个人邮箱的授权码)
set smtp-auth=login
set ssl-verify=ignore
set nss-config-dir=/etc/pki/nssdb/

cd /etc/pki/nssdb/
#生成证书

echo -n | openssl s_client -connect smtp.qq.com:465 | sed -ne ‘/-BEGIN CERTIFICATE-/,/-END CERTIFICATE-/p‘ > /etc/pki/nssdb/qq.crt 
certutil -A -n "GeoTrust SSL CA" -t "C,," -d /etc/pki/nssdb/ -i /etc/pki/nssdb/qq.crt 
certutil -A -n "GeoTrust Global CA" -t "C,," -d /etc/pki/nssdb/ -i /etc/pki/nssdb/qq.crt 
certutil -L -d /etc/pki/nssdb/ 
certutil -A -n "GeoTrust SSL CA - G3" -t "Pu,Pu,Pu" -d ./ -i qq.crt #认证

同样，认证完会返回如下提示：
Notice: Trust flag u is set automatically if the private key is present.

##测试
echo "this email come from centos 172.26.27.71"|mail -v -s "mysql check test" [email protected]

Part2 监控脚本准备

1. mysql监控脚本

大致想法：mysql监控脚本分别运行在两个实例上，如果当前实例宕机，则重启本机mysql服务，如果其他服务器上的mysql连接不上则邮件通知。

#!/bin/bash

notify_addr=‘[email protected],[email protected]‘
error_log="/opt/script/logs/check_mysql.err"

###定义一个简单判断mysql是否可用的函数
function excute_query {
    echo -e "`date "+%F  %H:%M:%S"`    -----checking mysql instance $1 by querying -----" >> ${error_log}
    /usr/local/mysql/bin/mysql -uroot -p88gongxiangMYSQL -h $1 --port 30468 -e "select 1;" 2>> ${error_log}
}

###定义无法执行查询，且mysql服务异常时的处理函数
function service_error {
    echo -e "`date "+%F  %H:%M:%S"`    -----mysql service error，notify manager now-----" >> ${error_log}
    systemctl restart mysql.service
    echo "$1 无法连接并被重启"|mail -s "MYSQL $1 实例正在被重启, 请及时登录查看状态！" ${notify_addr} 2>> ${error_log}
    echo -e "
---------------------------------------------------------
" >> ${error_log}
}

###定义无法执行查询,但mysql服务正常的处理函数
function query_error {
    echo -e "`date "+%F  %H:%M:%S"`    -----mysql instance $1 query error, retry after 30s-----" >> ${error_log}
    sleep 30
    excute_query $1
    if [ $? -ne 0 ];then
        echo -e "`date "+%F  %H:%M:%S"`    -----mysql instance $1 still can‘t execute query-----" >> ${error_log}
        echo "mysql isntance $1 is down"|mail -s "MYSQL $1 无法连接查询, 请及时处理！from(172.26.27.70)" ${notify_addr} 2>> ${error_log}
    else
        echo -e "`date "+%F  %H:%M:%S"`    -----mysql instance $1 query ok after 10s-----" >> ${error_log}
        echo -e "
---------------------------------------------------------
" >> ${error_log}
    fi
}

###监控本机mysql状态
excute_query 172.26.27.70
if [ $? -ne 0 ];then
     systemctl status mysql.service  &>/dev/null
    if [ $? -ne 0 ];then
        service_error 172.26.27.70
    else
        query_error 172.26.27.70
    fi
  else
     echo -e "
-----------mysql instance 172.26.27.70 is ok for query-------------
" >> ${error_log}
fi
###监控备机mysql状态
excute_query 172.26.27.71
if [ $? -ne 0 ];then
      query_error 172.26.27.71
    else
     echo -e "
-----------mysql instance 172.26.27.71 is ok for query-------------
" >> ${error_log}
fi

2. mongo监控脚本

大致思想：通过mongo命令登录或者mongostat判断节点是否正常运行。

notify_addr=‘[email protected],[email protected]‘
error_log="/opt/script/logs/check_mongo.err"

###定义一个简单判断mysql是否可用的函数
function connect_db {
    echo -e "`date "+%F  %H:%M:%S"`    -----checking mongo instance $1 by login -----" >> ${error_log}
    echo "db.serverStatus().mem" | /usr/local/mongodb/bin/mongo admin -uroot -p88gongxiangds --host $1 --port 20467  2>> ${error_log}
}

function replication_stat_query {
    echo -e "`date "+%F  %H:%M:%S"`    -----checking mongo instance $1 by mongostat -----" >> ${error_log}
    /usr/local/mongodb/bin/mongostat --uri=mongodb://suroot:[email protected]$1:20467/admin  2>> ${error_log}
}

###定义无法执行查询，且mysql服务异常时的处理函数
function service_error {
    echo -e "`date "+%F  %H:%M:%S"`    -----mongo service $1 error，notify manager now-----" >> ${error_log}
    ##/usr/local/mongo/bin/mongod -f /etc/mongo.conf --shutdown
    echo "$1 mongo连接失败，请及时处理"|mail -s "Mongo $1 实例无法连接, 请及时登录处理！from(172.26.27.70)" ${notify_addr} 2>> ${error_log}
    echo -e "
---------------------------------------------------------
" >> ${error_log}
}

###监控本机mongo node 状态
function monitor_node {
connect_db $1 
if [ $? -ne 0 ];then
    service_error $1
 #else
 #replication_stat_query $1
 #if [ $? -ne 0 ];then
 #service_error $1
 #else
 #echo -e "
-----------mongostat of node $1 is ok! -------------
" >> ${error_log}
           echo -e "
-----------mongo connection to  node $1 is ok! -------------
" >> ${error_log}
 #fi
fi
}
###监控本机mongo node 状态
monitor_node 172.26.27.70
monitor_node 172.26.27.71
monitor_node 172.26.27.72

3. redis监控脚本

大致思想： 通过redis-cli登录并检索clusterinfo是否enable来判断该节点及集群是否正常工作。

#!/bin/bash

notify_addr=‘[email protected],[email protected]‘
error_log="/opt/script/logs/check_redis.err"

###定义无法执行查询，且mysql服务异常时的处理函数
function service_error {
    echo -e "`date "+%F  %H:%M:%S"`    -----redis service $1:$2 error，notify manager now-----" >> ${error_log}
    ##/usr/local/mongo/bin/mongod -f /etc/mongo.conf --shutdown
    echo "$1 redis连接异常，请及时处理"|mail -s "Redis $1:$2 节点连接失败, 请及时登录处理！(from 172.26.27.70)" ${notify_addr} 2>> ${error_log}
    echo -e "
---------------------------------------------------------
" >> ${error_log}
}

###监控redis 状态
function monitor_node {
echo -e "`date "+%F  %H:%M:%S"`    -----checking mongo redis  $1:$2 by cli -----" >> ${error_log}
/usr/local/bin/redis-cli -h $1 -p $2 -a 88gongxiangrds info |grep  cluster_enabled
if [ $? -ne 0 ];then
    service_error $1 $2
    echo -e "
-----------redis connection to  node $1:$2 is ok! -------------
" >> ${error_log}
fi
}

###监控本机mongo node 状态
monitor_node 172.26.27.70 6239
monitor_node 172.26.27.70 6339
monitor_node 172.26.27.71 6239
monitor_node 172.26.27.71 6339
monitor_node 172.26.27.72 6239
monitor_node 172.26.27.72 6339

4. rabbitmq监控脚本

通过rabbitmqctl查看集群状态或者节点状态

#!/bin/bash

notify_addr=‘[email protected],[email protected]‘
error_log="/opt/script/logs/check_redis.err"

###定义无法执行查询，且mysql服务异常时的处理函数
function service_error {
    echo -e "`date "+%F  %H:%M:%S"`    -----rabbitmq service  error，notify manager now-----" >> ${error_log}
    #ps -ef | grep ^rabbitmq | awk ‘{print $2}‘ | xargs kill -9
    #service rabbitmq-server start
    echo "$1 rabbitmq服务异常, 请及时处理"|mail -s " $1 RabbitMQ服务异常, 请及时登录处理！（from $2）" ${notify_addr} 2>> ${error_log}
    echo -e "
---------------------------------------------------------
" >> ${error_log}
}

###监控rabbitmq 状态
function monitor_node {
echo -e "`date "+%F  %H:%M:%S"`    -----checking mongo redis  $1:$2 by cli -----" >> ${error_log}
#/usr/lib/rabbitmq/bin/rabbitmqctl  cluster_status |grep cluster_name 
/usr/sbin/rabbitmqctl cluster_status | grep cluster_name

if [ $? -ne 0 ];then
    service_error $1 $2
    echo -e "
-----------redis connection to  node $1:$2 is ok! -------------
" >> ${error_log}
fi
}

monitor_node 172.26.27.72 172.26.27.72

Part3 定时任务配置

crontab -e

*/1 * * * *    /opt/script/check_mysql.sh  > /opt/script/logs/cron_result.log 2>&1
*/3 * * * *    /opt/script/check_mongo.sh  > /opt/script/logs/cron_result.log 2>&1
*/5 * * * *    /opt/script/check_redis.sh  > /opt/script/logs/cron_result.log 2>&1
*/5 * * * *    /opt/script/check_rabbitmq.sh  > /opt/script/logs/cron_result.log 2>&1