[one-users] keepalived: problem implementing virtual ip

[Lim Kean Meng]

I think I found why not any ip just bind to eth0, the ip must be defined in context.sh.  In other words, the vm's network do not recognize any other ip except the one define in context.sh, see below:

Q: since I'll need 2 ip to bind to the same interface on eth0, how do I force the 2nd ip to be registered into context.sh?

root at one-dev04:/srv/cloud/one/var/262# more context.sh
# Context variables generated by OpenNebula
BROADCAST="10.4.104.255"
DNS="192.228.137.100"
FILES="/var/lib/one/vm-templates/ONE-centos/centos-init.sh"
GATEWAY="10.4.104.254"
HOSTNAME="CentOS-6.2-x64"
IP_PUBLIC="10.4.104.119"
NETMASK="255.255.255.0"
NETWORK="10.4.104.0"
TARGET="hdb"

Thanks and best regards.


Lim

From: Lim Kean Meng
Sent: Wednesday, 17 October, 2012 4:50 PM
To: 'users at lists.opennebula.org'
Subject: keepalived: problem implementing virtual ip

I want to setup a load balance to point to 2 VM in opennebula using keepalived, refer to ww.keepalived.org
I am having problem when implementing virtual ip or vip.
The vip which is actually an arbitrary ip in the same network  can bind to my VM eth0 (this is the master server) and if the master is down, the vip will float and bind to the standby VM eth0.

But the vip cannot be ping from other VM except the VM that is binding the vip as below.  Here my vip is 10.4.104.88 and my master and standby VM are 10.4.104.28 and 10.4.104.91 respectively, available below.

[root at DW-LB01-253 ~]# ping 10.4.104.88
PING 10.4.104.88 (10.4.104.88) 56(84) bytes of data.
64 bytes from 10.4.104.88: icmp_seq=1 ttl=64 time=0.127 ms
^C
--- 10.4.104.88 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 760ms
rtt min/avg/max/mdev = 0.127/0.127/0.127/0.000 ms

[root at DW-LB02-254 ~]# ping 10.4.104.88
PING 10.4.104.88 (10.4.104.88) 56(84) bytes of data.
^C
--- 10.4.104.88 ping statistics ---
1 packets transmitted, 0 received, 100% packet loss, time 804ms

[root at DW-LB01-253 ~]# ip add sh eth0
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 02:00:0a:04:68:1c brd ff:ff:ff:ff:ff:ff
    inet 10.4.104.28/24 brd 10.4.104.255 scope global eth0
    inet 10.4.104.88/32 scope global eth0
    inet6 fe80::aff:fe04:681c/64 scope link
       valid_lft forever preferred_lft forever

[root at DW-LB02-254 ~]# ip add sh eth0
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 02:00:0a:04:68:5b brd ff:ff:ff:ff:ff:ff
    inet 10.4.104.91/24 brd 10.4.104.255 scope global eth0
    inet6 fe80::aff:fe04:685b/64 scope link
       valid_lft forever preferred_lft forever


Thanks and best regards.


Lim

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20121018/3002ece1/attachment-0002.htm>