[Pacemaker] Private node IPs, public VIP and routing

[Dennis Jacobfeuerborn]

On 30.09.2013 13:24, Michael Schwartzkopff wrote:
> Am Montag, 30. September 2013, 13:11:28 schrieb Dennis Jacobfeuerborn:
>
>  > Hi,
>
>  > I am in the process of setting up a redundant NFS storage system and I'm
>
>  > wondering how to best manage the systems/IPs.
>
>  > The goal is that the nodes each get an IP in a management VLAN and the
>
>  > primary nodes gets a public VIP in a public facing VLAN in order to not
>
>  > have to allocate three public IPs for this one service.
>
>  > The issue is that in order to not run into routing issues the default
>
>  > route would have to be based in the network of the public VIP but since
>
>  > the VIP can move around and is not active on the secondary node I cannot
>
>  > set the default route there.
>
>  >
>
>  > Does anyone have an idea how to get around this problem or an
>
>  > alternative setup that accomplishes the same thing in a different way?
>
>  >
>
>  > Regards,
>
>  > Dennis
>
>  >
>
>  > _______________________________________________
>
>  > Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
>
>  > http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
>  >
>
>  > Project Home: http://www.clusterlabs.org
>
>  > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>
>  > Bugs: http://bugs.clusterlabs.org
>
> Hi,
>
> create a routing resource and group it together with the public VIP.

Hm, thanks for the pointer. While this would work with a regular static 
route the problem here is that the default route has to be replace i.e. 
the old one has to be removed first and then the second one added.
The one way I can see this working is by not configuring a default route 
at all on both nodes and then create two default route resources with a 
colocation rule that the private net default route is always started on 
the node where the public default route is *not* started.
I'm going to test this unless someone has a better idea how this could 
be improved upon.

Regards,
   Dennis