Ryan, I recall this lab.
You've changed the labels so they will be different. If you look at R8 and do "sh ip bgp vpnv4 vrf VPNB labels", look for the "In label /Out label" pushed onto the stack for route 200.0.0.7 (R7). If you're reachability is working and you've done the "neigh 200.0.0.9 next-hop-unchanged" on R1 & R9 (reverse to R1), then when you look at R2, run the command "sh ip bgp vpnv4 vrf VPNA labels". You should see the label being learned for 200.0.0.7 with a next hop of 200.0.0.8 R2 doesn't know how to get to R8 in terms of labels. (If you do "sh mpls forwarding-table vrf VPNA - you will see no label) R2: do a "sh ip cef vrf VPNA 200.0.0.7" You won't have a label path. No label for R8 (200.0.0.8). So, this is the reason why we need to issue the send-label cmd under the IPv4 address-family. ( I equate this send-label method similar to carving out the path like MPLS or sparse-mode in Multicast) On R2, then do "sh ip bgp ipv4 uni lab" and then you will see a label for 200.0.0.8. (Out label) On R2, sh ip cef vrf VPNA 200.0.0.7 and there will be 3 labels pushed onto the stack. Bottom label (right most label) is the VPN label, 2nd label (middle) tells us how to get to 200.0.0.8 and then the left most label tells us how to get to R5 R2: Now when you do a "sh mpls forw", you'll see the label pushed onto the stack for 200.0.0.5, the outgoing label. Go to R1: sh mpls forw - shows him popping the label off of the stack. Look at it all the way through the path, and you will now be able to get to R8-R7. Let me know if that helps with your customized labels. Thank you, Chris -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Ryan Jensen Sent: Sunday, January 26, 2014 10:08 PM To: [email protected] Subject: [OSL | CCIE_RS] Vol1. Lab 29 Verification. Hi All, I've completed this lab as it asks, but I am trying to more fully understand WHY it works. In AS 125, R1 is a route-reflector with R2 and R5 as clients. R5 connects to AS 689, and R2 connects to CE (R4). AS 689 is similar to AS 125 with R9 as RR for R6 and R9. R6 connects to R5 (AS 125) via EBGP and R8 connects to CE (R7). R1 and R9 have an eBGP peering as well. The Lab is relating to inter AS VPN I've been able to get R4 to ping R7 just fine. I understand how the labels are passed and whatnot, but what's confusing me is related to the connection between R1 and R2. For better understanding, on all the routers I've configured the label range to be x00-x50, I.e. R1 is 100 - 150, R2 is 200-250, etc. so i can get a feel for what labels i'm seeing. I've enabled 'debug mpls packet' on both routers R1 and R2 and issue the ping from R4 to R7. Here's what I see on R2 for the incoming packet from CE: *Jan 26 21:40:48.215: MPLS: Fa0/0: recvd: CoS=0, TTL=250, Label(s)=204 *Jan 26 21:40:48.215: MPLS: Se1/0.24: xmit: (no label) This seems OK, a bit odd that there's no label outbound no? Here's where it gets weird in my opinion, here's the output from R1: Jan 26 21:40:12.971: MPLS: Fa0/0.12: recvd: CoS=0, TTL=254, Label(s)=100/505/805 *Jan 26 21:40:12.975: MPLS: Fa0/0.15: xmit: CoS=0, TTL=253, Label(s)=505/805 Ok, so.... how was the packet xmit from R2 as untagged but arrive at R1 tagged with three labels? R1 and R2 are peered with BGP, OSPF, and LDP, there's nothing else between them. Without having to post all of the label bindings etc, anyone have any idea about this 'discrepencie'? _______________________________________________ Free CCIE R&S, Collaboration, Data Center, Wireless & Security Videos :: iPexpert on YouTube: www.youtube.com/ipexpertinc _______________________________________________ Free CCIE R&S, Collaboration, Data Center, Wireless & Security Videos :: iPexpert on YouTube: www.youtube.com/ipexpertinc
