ā04-07-2010 08:27 PM
Hi All,
I can't find much information or configuration examples about doing EoMPLS between different AS's.
I've got the basic setup:
PE1 -> P1 (7606) <--> P2 (7301) <--PE2
P1 & PE1 are in AS100
P2 & PE2 are in AS200
Vlan 37 originates on AS100 but I need PC's in AS200 to be in Vlan37 as well.
I have enabled BGP between the P routers.
For simplicity sake there is a /30 ethernet connection connecting the P routers.
Both AS's are running MPLS inside their core.
Could someone please point me in the right direction about what is required to get EoMPLS working between different AS's.
Thanks.
Andy
Solved! Go to Solution.
ā04-09-2010 01:30 AM
Hello Andy,
>>
My only question is that the chapter goes on to talk about seeing THREE labels in the vc details but in my lab I'm only seeing TWO labels (everything is working in my lab setup). Am I missing something here???
I haven't access to that book, however three labels make me think it was A Carrier supporting Carrier scenario using BGP with labels (the use of BGP with labels in CSC is possible)
I agree that in inter-AS scenario you should see only two labels.
The use of BGP next-hop-self can be of help in joining first level LSPs between loopbacks.
Hope to help
Giuseppe
ā04-08-2010 06:27 AM
Hello Andy,
EoMPLS use LDP targeted sessions between loopback ip addresses.
These IP addresses have to be known in the IP routing table by an IGP to create an LDP LSP with destination the remote loopback
the EoMPLS label becomes the inner label.
Your only option is to use BGP with labels
check if you can use neigh x.x.x.x send-label on your routers.
Once the LSP to the other device is built you should be fine.
see
http://www.cisco.com/en/US/docs/ios/12_0s/feature/guide/gsxnlbsp.html
recent new features for EoMPLS and VPLS include a new address-family for this purpose
Hope to help
Giuseppe
ā04-08-2010 05:47 PM
Thanks for the reply Giuseppe.
I was able to get more information by reading Layer 2 VPN Architectures by Wei Luo, Carlos Pignataro, Dmitry Bokotey, Anthony Chan. In there was a chapter dedicated to Pseudowire Emulation in Multi-AS Networks. And as you advised, the solution was to use the BGP "send-label" to redistribute the labels from one AS to another.
My only question is that the chapter goes on to talk about seeing THREE labels in the vc details but in my lab I'm only seeing TWO labels (everything is working in my lab setup). Am I missing something here???
From the book:
"Label 27 is the IGP label to reach ASBR1 that has the address 10.1.1.3, and label 32 is the BGP IPv4 label assigned by ASBR1 to reach PE4 that has the address 172.16.1.2"
PE1#show mpls l2transport vc 100 detail
Local interface: Et0/0.1 up, line protocol up, Eth VLAN 100 up
Destination address: 172.16.1.2, VC ID: 100, VC status: up
Preferred path: not configured Default path: active
Tunnel label: 32, next hop 10.23.12.2
Output interface: Et1/0, imposed label stack {27 32 16}
My Lab Setup:
test-mpls-cr#sh mpls l2transport vc detail
Local interface: Gi0/0.37 up, line protocol up, Eth VLAN 37 up
Destination address: 10.15.105.2, VC ID: 200, VC status: up
Output interface: Gi0/0.11, imposed label stack {7697 19}
Preferred path: not configured
Default path: active
Next hop: 203.10.110.195
Create time: 23:09:13, last status change time: 18:52:59
Signaling protocol: LDP, peer 10.15.105.2:0 up
MPLS VC labels: local 2851, remote 19
Group ID: local 0, remote 0
MTU: local 9000, remote 9000
Thanks.
Andy
ā04-09-2010 01:30 AM
Hello Andy,
>>
My only question is that the chapter goes on to talk about seeing THREE labels in the vc details but in my lab I'm only seeing TWO labels (everything is working in my lab setup). Am I missing something here???
I haven't access to that book, however three labels make me think it was A Carrier supporting Carrier scenario using BGP with labels (the use of BGP with labels in CSC is possible)
I agree that in inter-AS scenario you should see only two labels.
The use of BGP next-hop-self can be of help in joining first level LSPs between loopbacks.
Hope to help
Giuseppe
ā04-18-2010 10:28 PM
Extending on from this...what if the same vlan is being used at both ends? Say for example hosts in AS1 wants to access Vlan 37 resources in AS2, but AS1 is already using Vlan 37. Now you have a situation where you can't have matching Vlan ID's on the xconnect at both ends of the PE's.
[7301_PE] -> [7606_P] <--- (AS 1) EoMPLS (AS 2) ---> [7301_P] <--- [7301_PE]
I've read about this situation in another post below and you have to filter BPDU's to make it work.
EoMPLS with different vlans at the ends
https://supportforums.cisco.com/message/972918
Is filtering BPDU's the only way to make this work if the xconnect is using different Vlan ID's at each end??? I'm just worried about STP issues that can arise by filtering BPDU's that's all.
Thanks.
Andy
ā04-20-2010 12:03 AM
Hello Andy,
if the service is a single EoMPLS point-to-point service there are no STP issues (no redundant paths to deal with)
as in the linked thread there are some ways to deal with different Vlan-id at the two ends.
About filtering STP BPDUs I'm not sure if this happens automatically on EoMPLS (like automatic STP bpdu filter activated on 802.1Q tunnel interface).
As far as I know for example bpdu ethertype cannot be filtered on an ethertype ACL also on a C6500 (I had tried but the range of values of STP ethertype is not available for filtering)
Hope to help
Giuseppe
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide