It shouldn't do but I had some weird issues communicating between real
servers when this happened so I had to modify the IPAddr2 script to remove
the IP as there seems to be a bit of a regression with lvs_support=true
where it has to match the full IP and subnet to remove, but you need to
have the VIP as /32 on loopback and /24 on eth0
On 5 Dec 2013 08:52, "ESWAR RAO" <eswar7028@xxxxxxxxx> wrote:
> Hi all,
>
>
> I have a machine where the LVS runs with a VIP on eth0 and the 2 real
> backend servers have the same VIP in lo:0 interface. I have added
> arp_announce and arp_ignore in /proc/sys/net/ipv4/conf/*
>
> In the 3 node setup, whenever the LVS node fails , the LVS migrates to any
> one of the real nodes. But the VIP is also configured on the lo:0
> interface.
>
> Does it pose any problem to have same VIP on both eth0:0 and lo:0
> interfaces???
>
> Thanks
> Eswar
> _______________________________________________
> Please read the documentation before posting - it's available at:
> http://www.linuxvirtualserver.org/
>
> LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
> Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
> or go to http://lists.graemef.net/mailman/listinfo/lvs-users
>
_______________________________________________
Please read the documentation before posting - it's available at:
http://www.linuxvirtualserver.org/
LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
or go to http://lists.graemef.net/mailman/listinfo/lvs-users
|