L2 comunication between ASR Routers

[Nikolas2008]

Hello,

I realy need some help to understand an to make the configuration for trunk links and vlan access port on routers Cisco ASR 9001 and for ASR 902.

The thing is that I need to connect costumer devices on the ASR 902 tagged with vlans and connect this traffic through a trunk link to a ASR 9001 which have the connection to the Service provider devices. All traffic es L2 there is no IPs.

Thanks

[Peter L]

Hi

Check this link out on how to configure l2vpn in IOS-XR

http://www.cisco.com/c/en/us/support/docs/routers/asr-9000-series-aggregation-services-routers/116453-technote-ios-xr-l2vpn-00.html

Regards Peter

[xthuijs]

yeah XR due to the EVC model has no concept of "trunks" per-se, so you need to define EFP's or ether flow points aka l2transport subinterfaces for those vlans that you want to transport.

check the dekc from orland2013 from this query:

https://www.ciscolive.com/online/connect/search.ww#loadSearch-searchPhrase=2904&searchType=session&tc=0&sortBy=&p=&i%2810017%29=19409&i%2810017%29=16627&i%2810017%29=14187

it has some IOS ot XR comparison configurations that will probably help.

xander

[Nikolas2008]

Thanks for the reply.

I already did that l2 subinterface configuration, but I cant complete understand the rewrite command and how it works. 

If the connected devices are already sending traffic with 802.1q tags, do I need to remove those tags in order to send them throught the l2 subinterfaces ?

[xthuijs]

ah I see, I think I have a nice piece for you here.

it details out the precise func of that command.

if you have BVI's in your bridge domain you must pop the tags.

if you are doing l2 transparent services and no vlan translations or anything you *could* leave the tag as is, or there is no need to pop.

In the other hand, if the vlan is not service delimiting, you might as well pop it (to save some bytes) and push them at the remote end again. the rewrite operation doesnt cost you performance or anything.

cheers!

xander

[Nikolas2008]

Thanks Alexander for your help. One more thing, can you tell if the configuration is correct for this scenario?,

This is the topology.

The configuration apply is this: 

ASR920

interface GigabitEthernet0/0/1
 no ip address
 negotiation auto
 service instance trunk 1 ethernet
  encapsulation dot1q 51-52,76,99,150,451,1715,1717
  bridge-domain from-encapsulation
 !

interface GigabitEthernet0/0/5
 no ip address
 media-type auto-select
 negotiation auto
 service instance 51 ethernet
  encapsulation dot1q 51
  rewrite ingress tag pop 1 symmetric
  bridge-domain 51
 !
 service instance 52 ethernet
  encapsulation dot1q 52
  rewrite ingress tag pop 1 symmetric
  bridge-domain 52
 !

ASR9001

interface GigabitEthernet0/0/0/6
 description ##IFACE-to-ALU-QinQ##
!
interface GigabitEthernet0/0/0/6.76 l2transport
 encapsulation dot1q 100 second-dot1q 76
!
interface GigabitEthernet0/0/0/6.150 l2transport
 encapsulation dot1q 100 second-dot1q 150
!
interface GigabitEthernet0/0/0/6.451 l2transport
 encapsulation dot1q 100 second-dot1q 451
!
interface GigabitEthernet0/0/0/6.1715 l2transport
 encapsulation dot1q 200 second-dot1q 1715
!
interface GigabitEthernet0/0/0/6.1716 l2transport
 encapsulation dot1q 200 second-dot1q 1716
!
interface GigabitEthernet0/0/0/6.1717 l2transport
 encapsulation dot1q 200 second-dot1q 1717
!
interface GigabitEthernet0/0/0/10
 description ##IFACE-to-ALU 7450IMS##
!
interface GigabitEthernet0/0/0/10.51 l2transport
 encapsulation dot1q 51
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/10.52 l2transport
 encapsulation dot1q 52
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18
 description ##IFACE-TRUNK-to-ASR902##
!
interface GigabitEthernet0/0/0/18.1 l2transport
 encapsulation untagged
!
interface GigabitEthernet0/0/0/18.51 l2transport
 encapsulation dot1q 51
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18.52 l2transport
 encapsulation dot1q 52
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18.76 l2transport
 encapsulation dot1q 76
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18.99 l2transport
 encapsulation dot1q 99
!
interface GigabitEthernet0/0/0/18.150 l2transport
 encapsulation dot1q 150
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18.451 l2transport
 encapsulation dot1q 451
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18.1715 l2transport
 encapsulation dot1q 1715
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18.1716 l2transport
 encapsulation dot1q 1716
 rewrite ingress tag pop 1 symmetric
!
interface GigabitEthernet0/0/0/18.1717 l2transport
 encapsulation dot1q 1717
 rewrite ingress tag pop 1 symmetric
!

Thanks again. 

Best regards  

[xthuijs · Accepted solution]

Hi Nicolas, concept is fine, but I am missing the BD or XC configuration there.

One other thing to ntoe is that since you are popping the tag on the G0/0/0/18 side, you probably want to do that also on the Gig 0/0/0/6 side.

If you connect both EFP's together in an xcon like this:

l2vpn

xconnect group MYCONS

p2p first

int g 0/0/0/18.<x>

int g 0/0/0/6.<x>

for each EFP then you like them together, and what will happen is that ingress we pop the tag say 451 from GigabitEthernet0/0/0/18.451 and xcon that to GigabitEthernet0/0/0/6.451 and vbecause of the symmetric operation we push 100/451 on to the packet.

in reverse, if 100/451 comes in, we can pop that and xcon it to the g18 and apply 451 vlan dot1q single header only and you'd be good to go!!

BD when you need more then 2 endpoints.

XCON when you only need to connect 2 interfaces together (whether it be EFP, PW etc).

cheers!

xander