cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
376
Views
0
Helpful
4
Replies

OTV on ASR 1K

tahscolony
Level 1
Level 1

I have a question rolling around that I can't seem to pinpoint an answer to.  Lets say I have a 10 Gb L2 WAN connection that I want to run L3 over due to needing to separate Vlan and STP between sites.  However I still want to extend a couple Vlans to the new location.  SO I have a pair of ASr 1001-x with the 20 Gb license applied. One port connects to the 10GB WAN, with a subinterface with a /30, then a loopback interface for the BGP and OSPF peering, along with binding to the OTV.

Now the question I have that I can't seem to answer, the other 10 GB interface that connects to the core switch, which is doing L3 with OSPF as the routing protocol.  I want these switches to peer OSPF with the ASR's, and redistribute the OSPF through BGP over the WAN.

So how would this work with one 10GB interface connected to the switch from the ASR using OTV?  Is the interface a trunk port on the ASR to a trunk on the switch? The core switch on one side is Juniper, the other side has yet TBD.  Can I have a subinterface with a /30 on the same interface as the OTV?

4 Replies 4

Philip D'Ath
VIP Alumni
VIP Alumni

Have you considered using L2TP instead, and only extending the VLANs that you want to be layer 2?

L2TP just acts like a long piece of wire.  So the ASRs wont see or process any layer 3 on it.  So switches plugged into the L2TP ports will see each other as layer 2 adjacent.

Other VLANs, that don't have L2TP configure will just appear like a normal sub-interface which you can put an IP address on if you like.

We are looking at all aspects including L2TPv3, which I have deployed in the past, but not in this type of scenario where there will also be routing between core switches using the routers doing L2TPv3.

 

We want to create a BGP ring for redundancy, but may come upon an issue where servers being moved will not be able to have their IP's changed and need to be on the same network as other servers that are not being moved, yet we will also have new vlans and networks at the new location that will need to be routed back, also the ASR will be an Internet Router using VRF.  Right now there are many variables, too many options, and trying to narrow it down as to what technologies wont work so we can rule them out.

Ok,  for example, here is the OTV configuration from Cisco.

ASR-1
interface GigabitEthernet0/0/1
 no ip address
 negotiation auto
 cdp enable
 service instance 1 ethernet
  encapsulation untagged
  bridge-domain 1
 !
 service instance 50 ethernet
  encapsulation dot1q 100
  bridge-domain 200
 !
 service instance 51 ethernet
  encapsulation dot1q 101
  bridge-domain 201

So now I can add
interface GigabitEthernet0/0/1.1
ip address 1.1.1.1 255.255.255.252

If I have 1.1.1.2 255.255.255.252 on interface vlan 1 on the switch, and setup OSPF using a loopback as the source, this should work?

What your doing really screams LISP to me.  The boring marketing bit is here:

http://lisp.cisco.com/

Here is an actual configuration example showing the same subnet being used at two different sites at the same time.

http://www.cisco.com/c/en/us/td/docs/solutions/Enterprise/Data_Center/DCI/5-0/LISPmobility/DCI_LISP_Host_Mobility/LISPmobile_5.html

LISP is kinda a competitor to OTV.  However it can also be extended to WANs (optionally), and integrates both layer 2 and layer 3 at the same time. it can also be used on much smaller hardware.  So you can extend it out to a little Cisco 890 series at a branch.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Review Cisco Networking products for a $25 gift card