Keepalived and Mysql Replication

Jerim79 · 03-07-2012, 01:01 AM

I have two servers with mysql installed. MySQL replication works fine until I turn on keepalived on the slave. When I do, I get this message:

[ERROR] Slave I/O: error reconnecting to master 'repl@10.89.96.144:3306' - retry-time: 60 retries: 86400, Error_code: 2013

Here is the keepalived.conf file on the slave:

vrrp_instance VI_1 {
state SLAVE
interface eth0
virtual_router_id 11
priority 100

virtual_ipaddress {
10.89.96.56/24

}
notify_master "/etc/keepalived/notify.sh del 10.89.96.56"
notify_backup "/etc/keepalived/notify.sh add 10.89.96.56"
notify_fault "/etc/keepalived/notify.sh add 10.89.96.56"

}

virtual_server 10.89.96.56 80 {
delay_loop 30
lb_algo rr
lb_kind DR
persistence_timeout 50
protocol TCP

real_server 10.89.96.144 80 {
weight 2
HTTP_GET {
url {
path /index.php
digest 82520e8260668136231dc2a41d81309d
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 2
}

}

real_server 10.89.96.127 80 {
weight 1
HTTP_GET {
url {
path /index.php
digest cd893ebd870b545d3d6cdcbb25d4510b
}
connect_timeout 3
nb_get_retry 3
delay_before_retry 2
}

}

}

mulyadi.santosa · 03-08-2012, 11:26 AM

Hi ...

It's been a long time ago since I worked with keepalived, but reading your described symptom, I have a guess that somehow keepalived "shutdown" the network connection to your mysql slave.

It could be due to something that block the http probing, like described in your keepalived config.

BTW, about Mysql error code, I got hints from this URL:
http://serverfault.com/questions/295...rror-2013-mean

hope it helps...