Comments on: IP FailOver Setup Using KeepAlived on CentOS & Red Hat https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/ How to guide for System Administrator's and Developers Mon, 21 Oct 2019 06:49:18 +0000 hourly 1 https://wordpress.org/?v=6.1.1 By: Kavshal https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-43316 Mon, 21 Oct 2019 06:49:18 +0000 https://tecadmin.net/?p=630#comment-43316 Hi Rahul,

I have below setup for mysql:
Machine #A – 192.168.10.1
Machine #B – 192.168.10.2
VIP between Machine #A and Machine #B is 192.168.10.3

Let say keeapalived is master in Machine #A, possible the application not to use virtual ip to connect to the destination. Sample example is connecting to mysql:
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
tcp 0 0 192.168.10.3:32444 192.168.10.3:3306 ESTABLISHED
tcp 0 0 192.168.10.3:32442 192.168.10.3:3306 ESTABLISHED

What I’m expecting is as below:
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
tcp 0 0 192.168.10.1:32444 192.168.10.3:3306 ESTABLISHED
tcp 0 0 192.168.10.1:32442 192.168.10.3:3306 ESTABLISHED

Anyway I can achieve above?

]]>
By: Rahul https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-43176 Mon, 14 Oct 2019 07:26:07 +0000 https://tecadmin.net/?p=630#comment-43176 Yes

]]>
By: saravanakumar https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-43134 Fri, 11 Oct 2019 06:53:21 +0000 https://tecadmin.net/?p=630#comment-43134 waiting for your reply

]]>
By: saravana kumar https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-43067 Tue, 08 Oct 2019 09:58:42 +0000 https://tecadmin.net/?p=630#comment-43067 will it work in different network ?

]]>
By: Rahul https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-42696 Wed, 18 Sep 2019 04:48:49 +0000 https://tecadmin.net/?p=630#comment-42696 Yes, you can use public IP.

]]>
By: saravanakumar https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-42679 Tue, 17 Sep 2019 14:01:14 +0000 https://tecadmin.net/?p=630#comment-42679 shall i give public IP (142.93.12.140) instead of Local IP(192.168.10.121)..??

]]>
By: patricklbs https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-5200 Thu, 05 Apr 2018 14:01:48 +0000 https://tecadmin.net/?p=630#comment-5200 Why install gcc + kernel headers/devel ?

]]>
By: Sam https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-3984 Thu, 20 Jul 2017 12:27:08 +0000 https://tecadmin.net/?p=630#comment-3984 Thank you very much for this tutorial. I am a beginner and I found it really easy to follow. Cheers!

]]>
By: Ram J https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-173 Wed, 16 Sep 2015 14:36:58 +0000 https://tecadmin.net/?p=630#comment-173 Hi,
First of all this is a post with good explanation. I have a couple of questions regarding sync between servers. As we all know failover/redundancy servers needs to be synchronised to have consistent data across all the servers. One of the site recommended to run rsync to sync data between master and secondary servers periodically. At the worst case consistent data may not be available on a particular server at that point of time when sync hasn’t happened yet. So I was wondering is it possible to feed data from the firewall itself to multiple systems in parallel so that there wont be any sync issues? Kindly let me know. Thanks

]]>
By: rcb https://tecadmin.net/ip-failover-setup-using-keepalived-on-centos-redhat/#comment-172 Tue, 23 Dec 2014 11:13:45 +0000 https://tecadmin.net/?p=630#comment-172 Hi Jaffar,
vrrp_script is a script which will be running every 2 s (interval 2), so i dont think that you want that every 2s “killall -0 haproxy” will be executed. Thus i think you should run a command that check the state of your proxy like ps or a special command of your proxy. Anyway, i think that the fail over happens because if the secondary server is the first which runs keepalived, he will runs “killall -0 haproxy”, and increment the priority by 2 (weight 2), so the priority becomes 102 > 101 (the priority of the and increment the priority by 2 (weight 2)primary server) => FAILOVER, after that as the same way, the primary server runs “killall -0 haproxy” and increment the priority by 2 (weight 2), so the priority becomes 103 > 102 => FAILOVER ….

]]>