Home > OS >  About the high availability of keepalived VIP address migration issues
About the high availability of keepalived VIP address migration issues

Time:09-18

Everybody is good, when I was made a high availability keepalived met a strange problem: configuration, simulation stop keepalived, VIP address can properly from the master server to server, but I repeat the test several times, more open after keepalived shut down more than a few times, it is out of question, VIP is a fixed die on the primary server, cannot migrate to prepare on the server, the configuration file is as follows:
The primary server:
/root @ lb01 ~ # cat/etc/keepalived/keepalived conf
! The Configuration File for keepalived

Global_defs {
Router_id lb01
}


Vrrp_instance keep {
State MASTER
Interface eth0
Virtual_router_id 51
The priority of 150
Advert_int 1
The authentication {
The auth_type PASS
Auth_pass 1111
}
Virtual_ipaddress {
192.168.22.3/24 dev eth0 label eth0:3
}

}

For the server configuration:
/root @ lb02 ~ # cat/etc/keepalived/keepalived conf
! The Configuration File for keepalived

Global_defs {
Router_id lb02
}

Vrrp_instance keep {
The state BACKUP
Interface eth0
Virtual_router_id 51
The priority of 100
Advert_int 1
The authentication {
The auth_type PASS
Auth_pass 1111
}
Virtual_ipaddress {
192.168.22.3/24 dev eth0 label eth0:3
}
}

CodePudding user response:

BACKUP the priority on this machine to 130 or 140 line, the reason is the MASTER and BACKUP of the difference of the priority of the less than a certain WEIGHT values

CodePudding user response:

reference 1st floor crynono response:
BACKUP the priority on this machine to 130 or 140 line, the reason is that the MASTER and BACKUP of the difference of the priority of the less than a certain WEIGHT values

As if also not line, tried several Numbers, is still the same result

CodePudding user response:

To answer the problem solved, oneself, let people encounter the same problems can reference, the initial configuration no problem, just stop test keepalived services, were not completely stopped, as if systemctl stop keepalived is just stop the main process, has not stopped the child, should use systemctl kill keepalived, it normal switching the VIP address
  • Related