[j-nsp] Strange behaviour MX and VPLS

Martin Levin martin.levin at molndal.se
Fri Oct 7 10:10:53 EDT 2011


Hi,
 
All is happening inside the MX480. No VPLS traffic ever leaves the
router.
 
A core interface is an interface that has 
 
family bridge {
    core-facing;
}
 
enabled on it.
 
Example configuration.
 
VPLS routing-instance
 
duc {
    instance-type vpls;
    vlan-id 41;
    interface ae0.12;
    interface ae.1744
    no-local-switching;
    protocols {
        vpls {
            no-tunnel-services;
        }
    }
}
Interface ae0:
 
unit 12 {
    description DUC;
    encapsulation vlan-vpls;
    vlan-id 41;
    family vpls {
        core-facing;
    }
}
Interface ae1:
 
 
unit 1744 {
    description sthu_plan1_duc_nya;
    encapsulation vlan-vpls;
    vlan-tags outer 551 inner 14;
    family vpls {
        core-facing;
    }
}
In this case the equipment connected to the two interfaces won't be
able to communicate with each other.
 
//Martin

 
 
---
Martin Levin
IT-strategy & planning
Mölndals stad
>>> 


Från:David Ball <davidtball at gmail.com>
Till:Martin Levin <martin.levin at molndal.se>
Kopia:<juniper-nsp at puck.nether.net>
Datum:2011-10-06 21:10
Ärende:Re: [j-nsp] Strange behaviour MX and VPLS
  Is this all happening within the single MX480 router?  Can you
differentiate between a core and a normal interface ?  Is this LDP- or
BGP-signalled VPLS ?  Any configs you can share ?

David

On 6 October 2011 10:49, Martin Levin <martin.levin at molndal.se> wrote:
> Hi,
>
> At https://www.sugarsync.com/pf/D6892768_9896580_18584 you'll find a
> logical map of a vpls domain connections.
>
> For background information: We have devices in our network that need
to
> be able to communicate with servers, and sometimes with other devices
as
> well. To limit mac-adress distribution in our network we want to be
able
> to run our vpls instances in no-local-switching mode.
>
> If we look at the picture the expected behaviour is the following.
>
> PC1 + PC2 should be able to communicate with everything and
everything
> should be able to communicate with them.
> PC3 + PC4 should be able to communicate with PC1 + PC2 but not with
> each other.
>
> Actual behaviour is:
>
> One core (PC1) interface (probably the first in the configuration)
is
> able to communicate with everything except the other core interface.
> Nothing is able to communicate with the other core interface (PC2).
> Normal interfaces (PC3, PC4) are able to communicate with core
(PC1),
> but not with each other.
>
> Nothing in the manual gives any indication that only ONE core-facing
> interface is allowed.
>
> Any help to reach the expected behaviour would be greatly
appreciated.
>
> Sincerely,
> Martin
> City of Molndal
>
>
> ---
> Martin Levin
> IT-strategy & planning
> Mölndals stad
> _______________________________________________
> juniper-nsp mailing list juniper-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/juniper-nsp


More information about the juniper-nsp mailing list