LVS Health Check

Environment: centos7.1

Brief introduction:

When the script detects that an RS HTTP service is dropped, RS is automatically removed from the LVS, and when all RS HTTP services are dropped, all RS is removed in the LVS and the HTTP service on the LVS Scheduler is added to the LVS as an alarm page.

When the script detects that an RS HTTP service is active, it is automatically added to the LVS and deleted if there is a LVS in the scheduler itself.

The script is tested by me and fully meets the requirements.

Here is the script:

#!/bin/bash#rs= ("152.168.1.12"   "152.168.1.13") vip= "152.168.1.10" port=80logfile= "/usr/local/scripts/ Lvs.log "function check_alldown {    #有一个rs主机能访问, that means it's not all gone    # Check to an RS host to survive and exit the check     #如果全部rs不能访问, indicating that the host has lost    for www in  ' echo ${ Rs[*]} '    do      curl --connect-timeout 1 http:// $www  &> /dev/null      if [ $? -eq 0 ]       then          echo  0           exit 0        fi         done   echo 100 } function lvs_add {   ipvsadm -a -t  $VIP: $port  -r $1    echo  "add  Rs host:$1 to lvs "}function lvs_rm {   ipvsadm -d -t $ VIP: $port  -r $1   echo  "Remove rs host:$1 to lvs"}function  lvs_local {    #如果全部rs主机掉线 and add it     #如果可以访问一个rs主机 without 127.0.0.1 in LVs, And the LVS have 127.0.0.1 to delete it    all_down= ' Check_alldown '    rip=$ (ipvsadm -l -n  | gawk  '/127.0.0.1/')    if [  $all _down -eq 100 ]    then       if [  "$rip"  =  " ]         then               echo  "' date +%f:%h-%m-%s '  all rs host is down!"  >>  $logfile            lvs_add  " 127.0.0.1 "       fi   else       if [  $all _ down -eq 0 ] && [ !  "$rip"  =  ""  ]          then           echo   "' date +%f:%h-%m-%s '  one rs host is up,remove local rs host !"  >>  $logfile            lvs_rm  " 127.0.0.1 "       fi   fi}function lvs_rs {     #如果可以访问一个rs主机, and the LVS does not add it     #如果不能访问一个rs主机, and it removes it in LVs    lvs_local    for www in  ' echo ${rs[*]} '    do       rip=$ (ipvsadm -l -n | gawk  "/$www/")        curl --connect-timeout 1 http://$www  &> /dev/null      if [  $? -eq 0 ]      then           if [  "$rip"  =  " ]           then              echo   "' date +%f:%h-%m-%s '  rs host: $www  is up!"  >>  $logfile                lvs_add  "$www"           fi       else          if [ !  "$rip"  =  ""  ]          then               echo  " ' date +%f:%h-%m-%s '  rs host: $www  is down! "  >>  $logfile                lvs_rm  "$www"           fi       fi   done}function lvs_monitor {   while true    do     echo  "check lvs rs health!"      lvs_rs     sleep 1   done}lvs_ Monitor

Complete.

This article is from the "Super_color" blog, make sure to keep this source http://small.blog.51cto.com/259970/1728082

LVS Health Check