LB层到Real Server之间访问请求的响应时间及HTTP状态码监控及报警设置

为了监控到各业务的访问质量，基于LB层的Nginx日志，实现LB层到Real Server之间访问请求的响应时间（即upstream_response_time）及HTTP状态码（即upstream_status）的监控及报警。操作记录如下：

基本信息：
负载均衡采用的是Nginx+Keeplived
负载域名：bs7001.kevin-inc.com （有很多负载域名，这里用该域名作为示例）
日志：bs7001.kevin-inc.com-access.log

1）LB层Nginx的log_format日志格式的设置（可以参考：http://www.cnblogs.com/kevingrace/p/5893499.html）
[root@inner-lb01 ~]# cat /data/nginx/conf/nginx.conf
......
######
    ## set access log format
    ######
    log_format  main  '$remote_addr $remote_user [$time_local] "$request" '
                  '$status $body_bytes_sent "$http_referer" '
                  '$http_user_agent $http_x_forwarded_for $request_time $upstream_response_time $upstream_addr $upstream_status';

    #######
.....

2）监控及报警脚本设置
日志路径
[root@inner-lb01 ~]# ll /data/nginx/logs/bs7001.kevin-inc.com-access.log
-rw-r--r-- 1 root root 0 12月 13 17:00 /data/nginx/logs/bs7001.kevin-inc.com-access.log

sendemail安装配置（安装可参考：http://www.cnblogs.com/kevingrace/p/5961861.html）
[root@inner-lb01 ~]# cat /opt/sendemail.sh        //该脚本可直接拿过来使用
#!/bin/bash
# Filename: SendEmail.sh
# Notes: 使用sendEmail
#
# 脚本的日志文件
LOGFILE="/tmp/Email.log"
:>"$LOGFILE"
exec 1>"$LOGFILE"
exec 2>&1
SMTP_server='smtp.kevin.com'
username='[e-mail]'
password='notice@123'
from_email_address='[e-mail]'
to_email_address="$1"
message_subject_utf8="$2"
message_body_utf8="$3"
# 转换邮件标题为GB2312，解决邮件标题含有中文，收到邮件显示乱码的问题。
message_subject_gb2312=`iconv -t GB2312 -f UTF-8 << EOF
$message_subject_utf8
EOF`
[ $? -eq 0 ] && message_subject="$message_subject_gb2312" || message_subject="$message_subject_utf8"
# 转换邮件内容为GB2312，解决收到邮件内容乱码
message_body_gb2312=`iconv -t GB2312 -f UTF-8 << EOF
$message_body_utf8
EOF`
[ $? -eq 0 ] && message_body="$message_body_gb2312" || message_body="$message_body_utf8"
# 发送邮件
sendEmail='/usr/local/bin/sendEmail'
set -x
$sendEmail -s "$SMTP_server" -xu "$username" -xp "$password" -f "$from_email_address" -t "$to_email_address" -u "$message_subject" -m "$message_body" -o message-content-type=text -o message-charset=gb2312

[root@inner-lb01 ~]# cd /opt/lb_log_monit.sh/
[root@inner-lb01 lb_log_monit.sh]# ll
总用量 12
-rwxr-xr-x 1 root root 1180 2月   1 13:03 bs7001_request_status_monit.sh
-rwxr-xr-x 1 root root  821 2月   1 11:20 bs7001_request_time_monit_request.sh
-rwxr-xr-x 1 root root  559 2月   1 13:01 bs7001_request_time_monit.sh

访问请求的响应时间监控报警脚本（下面脚本中取日志文件中的第3、10列以及倒数第1、2、3列）
[root@inner-lb01 lb_log_monit.sh]# cat bs7001_request_time_monit.sh
#!/bin/bash
/usr/bin/tail -1000 /data/nginx/logs/bs7001.kevin-inc.com-access.log|awk '{print $3,$10,$(NF-2),$(NF-1),$(NF)}' > /root/lb_log_check/bs7001.kevin-inc.com-check.log

for i in `awk '{print $3}' /root/lb_log_check/bs7001.kevin-inc.com-check.log`
do
  a=$(printf "%f" `echo ${i}*1000|bc`|awk -F"." '{print $1}')
  b=$(printf "%f" `echo 1*1000|bc`|awk -F"." '{print $1}')

  if [ $a -ge $b ];then
     cat /root/lb_log_check/bs7001.kevin-inc.com-check.log |grep $i
  else
     echo "it is ok" >/dev/null 2>&1
  fi
done

[root@inner-lb01 lb_log_monit.sh]# cat bs7001_request_time_monit_request.sh
#!/bin.bash
/bin/bash -x /opt/lb_log_monit.sh/bs7001_request_time_monit.sh > /root/lb_log_check/bs7001.kevin-inc.com_request_time.log

NUM=`cat /root/lb_log_check/bs7001.kevin-inc.com_request_time.log|wc -l`
if [ $NUM != 0 ];then
  /bin/bash /opt/sendemail.sh [e-mail] "从LB层访问bs7001.kevin-inc.com请求的响应时间" "响应时间已超过1秒钟!\n具体情况如下:\n`cat /root/lb_log_check/bs7001.kevin-inc.com_request_time.log`"
  /bin/bash /opt/sendemail.sh [e-mail] "从LB层访问bs7001.kevin-inc.com请求的响应时间" "响应时间已超过1秒钟!\n具体情况如下:\n`cat /root/lb_log_check/bs7001.kevin-inc.com_request_time.log`"
else
  echo "从LB层访问bs7001.kevin-inc.com请求的响应正常"
fi

[root@inner-lb01 lb_log_monit.sh]# ll /root/lb_log_check/
总用量 152
-rw-r--r-- 1 root root 147766 2月   1 15:00 bs7001.kevin-inc.com-check.log
-rw-r--r-- 1 root root    216 2月   1 15:00 bs7001.kevin-inc.com_request_time.log

访问的HTTP状态码监控报警脚本（500,502,503,504的状态码进行报警）
[root@inner-lb01 lb_log_monit.sh]# cat bs7001_request_status_monit.sh
#!/bin/bash
/usr/bin/tail -1000 /data/nginx/logs/bs7001.kevin-inc.com-access.log|awk '{print $3,$10,$(NF-2),$(NF-1),$(NF)}' > /root/lb_log_check/bs7001.kevin-inc.com-check.log

for i in `awk '{print $5}' /root/lb_log_check/bs7001.kevin-inc.com-check.log|sort|uniq`
do
  if [ ${i} = 500  ];then
    /bin/bash /opt/sendemail.sh [e-mail] "从LB层访问bs7001.kevin-inc.com请求的HTTP状态返回码" "HTTP状态返回码:500\n具体情况如下:\n`cat /root/lb_log_check/bs7001.kevin-inc.com-check.log |grep ${i}`"
  elif [ ${i} = 502  ];then
    /bin/bash /opt/sendemail.sh [e-mail] "从LB层访问bs7001.kevin-inc.com请求的HTTP状态返回码" "HTTP状态返回码:502\n具体情况如下:\n`cat /root/lb_log_check/bs7001.kevin-inc.com-check.log |grep ${i}`"
  elif [ ${i} = 503  ];then
    /bin/bash /opt/sendemail.sh [e-mail] "从LB层访问bs7001.kevin-inc.com请求的HTTP状态返回码" "HTTP状态返回码:503\n具体情况如下:\n`cat /root/lb_log_check/bs7001.kevin-inc.com-check.log |grep ${i}`"
  else
     echo "it is ok"
  fi
done

3）结合crontab进行定时监控
[root@inner-lb01 lb_log_monit.sh]# crontab -l
#LB到后端服务器之间访问各系统业务的请求响应时间和http状态码监控
*/2 * * * * /bin/bash -x /opt/lb_log_monit.sh/bs7001_request_time_monit_request.sh >/dev/null 2>&1
*/2 * * * * /bin/bash -x /opt/lb_log_monit.sh/bs7001_request_status_monit.sh >/dev/null 2>&1

取对应log文件中的第3、10以及倒数第1、2、3列内容
[root@inner-lb01 lb_log_monit.sh]# /usr/bin/tail -10 /data/nginx/logs/bs7001.kevin-inc.com-access.log|awk '{print $3,$10,$(NF-2),$(NF-1),$(NF)}'
[01/Feb/2018:15:05:41 "http://bs7001.kevin-inc.com/I8QW/dataimport/qw/found_analysis_imp8.jsp" 0.002 192.168.1.22:7001 304
[01/Feb/2018:15:05:41 "http://bs7001.kevin-inc.com/I8QW/dataimport/qw/found_analysis_imp8.jsp" 0.001 192.168.1.22:7001 304
[01/Feb/2018:15:05:41 "http://bs7001.kevin-inc.com/I8QW/dataimport/qw/found_analysis_imp8.jsp" 0.002 192.168.1.22:7001 304
[01/Feb/2018:15:05:41 "http://bs7001.kevin-inc.com/I8QW/dataimport/qw/found_analysis_imp8.jsp" 0.001 192.168.1.22:7001 304
[01/Feb/2018:15:05:41 "http://bs7001.kevin-inc.com/I8QW/dataimport/qw/found_analysis_imp8.jsp" 0.001 192.168.1.22:7001 304
[01/Feb/2018:15:05:41 "http://bs7001.kevin-inc.com/I8QW/dataimport/qw/found_analysis_imp8.jsp" 0.002 192.168.1.22:7001 304
[01/Feb/2018:15:06:02 "http://bs7001.kevin-inc.com/portal/main_new.do" 0.006 192.168.1.21:7001 200
[01/Feb/2018:15:07:12 "http://bs7001.kevin-inc.com/portal/main_new.do" 0.003 192.168.1.22:7001 200
[01/Feb/2018:15:07:51 "http://bs7001.kevin-inc.com/portal/main_new.do" 0.003 192.168.1.21:7001 200
[01/Feb/2018:15:07:57 "http://bs7001.kevin-inc.com/portal/main_new.do" 0.007 192.168.1.22:7001 200