[c-nsp] Readvertising strangeness.

[Gustav Ulander]

Hello.

Got a tip of list to try and change the label mode which I did. 
Configuring the PE router with the IPV4 peer with per-ce and a reboot solved it for CSR1000v. :)

//Gustav 


-----Original Message-----
From: cisco-nsp [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Gustav Ulander
Sent: den 19 november 2015 08:29
To: 'cisco-nsp at puck.nether.net' <cisco-nsp at puck.nether.net>
Subject: Re: [c-nsp] Readvertising strangeness.

Hello.

So a little update and background. 
What iam trying to accomplish is to change the AS number in half the network since we are going to connect 2 different AS numbers. 
The idea was to connect the networks via IGP in the core nodes together with LDP to get transport labels across and then just peer between the Routereflectors to get vpnv4 advertisements across. With this setup we would be able to provision Customer IP-VPN across the 2  networks without having to stitch the VRFs or having all the traffic move between PE routers as in inter-as option 2 (which seems to be the closest to what iam thinking).

//Gustav


-----Original Message-----
From: cisco-nsp [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Gustav Ulander
Sent: den 17 november 2015 11:04
To: 'cisco-nsp at puck.nether.net'
Subject: Re: [c-nsp] Readvertising strangeness.

Hello people.

I have some strange things going on here.
Iam doing  some test configurations in the lab and are trying to see how 2 AS numbers behave when they peer via their respective route reflectors.
I am using CSR1000v for this lab as a start.
I have 2 AS numbers to start with. Every AS has 1 PE router, 1 P router and 1 route reflector.
The PE and RR are doing IBGP and MPLS. The P router is doing IGP (OSPF) and MPLS. The 2 AS numbers are connected via IGP between P routers.
The two route reflectors are doing EBGP between themselves with next-hop unchanged to allow the PE routers to send traffic directly towards each other.
Everything Is working as expected and Iam able to import and export route targets across ASes when I have locally attached interfaces.
The strange part arises when I add EBGP links to the PE routers.
I can have a VRF that has a loopback in each PE router ping works perfectly between them and traceroute says that iam using the label switched path via P routers.
However If I introduce a EBGP link in the VRF the router that has the link can no longer ping the other PE router. The PE router that dosent have the link can ping the first PE router.
So it seems that when I introduce a EBGP link in the vrf the forwarding stops working.
The routes from the other is still in the table and labels are assigned and so forth. Anyone have any input?


//Gustav
_______________________________________________
cisco-nsp mailing list  cisco-nsp at puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
_______________________________________________
cisco-nsp mailing list  cisco-nsp at puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/