[DRBD-user] drbd8 - multiple routes between nodes

[Guy]

Hi Leandro,

I'm running 2 nodes with a primary/secondary & secondary/primary
setup. So one primary partition per node under normal running
conditions. The partitions are mounted as nfs exports. If one node
goes down it's fairly easy to resync back and then switch back to a
primary partition per node once the downed node is returned. If one of
the nodes loses connection I land up with split brain. Normal fencing
rules seem to favour the old primary and outdate the secondary so it
can't switch to primary which isn't what I need. Since the servers
will have 3 interfaces (1 crossover, 1 internal network, 1 external
IP) it would have been a lot easier to keep drbd communicating if
multiple routes were possible.

Thanks
Guy

On 19/03/2008, Leandro Dardini <ldardini at tiscali.it> wrote:
> Why don't you tell us the reason to have two address, maybe we can give
>  you a better solution.
>
>  Leandro
>
>  Grégory Steulet ha scritto:
>
> > Hello Guy,
>  >
>  > I was also looking for such a possibility. Unfortunatelty for the moment the only way to do that is bonding your eth devices. I'm testing this configuration on VmWare with an SLES10, for the moment it looks good.
>  >
>  > Greg
>  >
>  >
>  >
>  >
>  > -----Original Message-----
>  > From: drbd-user-bounces at lists.linbit.com on behalf of Guy
>  > Sent: Wed 3/19/2008 11:28 AM
>  > To: drbd-user at lists.linbit.com
>  > Subject: [DRBD-user] drbd8 - multiple routes between nodes
>  >
>  > Hi,
>  >
>  > I've just started using drbd8 with heartbeat v2 + CRM with the hopes
>  > of using it in a production environment relatively soon. Testing so
>  > far has been great, but is it possible to use more than one address
>  > for a device?
>  >
>  > Something like:
>  >   on drbd1 {
>  >     device     /dev/stor1;
>  >     disk       /dev/hdc4;
>  >     address    10.0.0.11:7788;
>  >     address    217.13.49.111:7788;
>  >     meta-disk  /dev/hdc2[0];
>  >   }
>  >
>  >   on drbd2 {
>  >     device     /dev/stor1;
>  >     disk       /dev/hdc4;
>  >     address    10.0.0.12:7788;
>  >     address    217.13.49.112:7788;
>  >     meta-disk  /dev/hdc2[0];
>  >   }
>  >
>  > Or perhaps have failover ips as the machines will two interfaces set
>  > up on different networks? If drbd can use both networks it would
>  > provide an extra level of redundancy to the connectivity between the
>  > nodes.
>  >
>  > Thanks
>  > Guy
>  >
>  >
>
>


-- 
Don't just do something...sit there!