Cisco SD-WAN 017 - Service VPN1 BGP Peering Setup and Route Propagation

[@ali_HA24]

EXCELLENT TSHOOT SKILLS, GOT FULL UNDERSTANDING ON THE LOGIC AND REALLY HELPED THE WAY YOU EXPLAINED THE WHOLE CONCEPT. BEST SD-WAN SERIES SO FAR ON KZbin, LOOKING FORWARD TO THE NEXT VIDEO. PLEASE ALSO CONFIRM IF YOU WILL' BE MAKING VIDEO ABOUT SERVICE CHAINING I.E. WHERE TRAFFIC FROM DIFFERENT SITES HAVE TO GO THROUGH FIREWALL BEFORE THEY CAN TALK TO OTHER SITES..

[@RobRikerTechChannel]

Its on the list, working on a VPN segmentation video now

[@jonallan121]

Ahhh man! From the moment I saw you redistribute the Internal OSPF I was trying to do ‘time travel KZbin telepathy’ to tell you to “stoooop Rob, we will lose the loopbacks for the BGP Peering!!!” - Haha. I done the exact same thing a while ago (in a live environment) and believe me, I’ve never let it catch me out again! Another top video mate! It’s a very clever use of iBGP!

[@RobRikerTechChannel]

Yeah, sometimes bloopers make the videos more interesting, teachable/learning moments hey!

[@jonallan121]

Rob Riker's Tech Channel couldn’t agree more ! I wish I had saw a video with it before that dreaded live environment event ! You may potentially save someone’s job or at the very least their sanity by leaving these things in!

[@jonallan121]

The Dark Knight It’s nothing major dude, Rob resolved it and done some awesome Educational troubleshooting. Rob was advertising the LOOPBACKS for the BGP peering via OSPF but then had BGP being preferred over OSPF which included the loopbacks. He intended to only increase OSPF external routed admin distance so it would not affect the internal OSPF loopback routes but he overlooked it, then realised near the end of the vid when he was having issues. It’s an easy error to make, one which we will all come across at some point and can often forget about so it’s good that Rob showed it in this video.

[@Valtrom87]

Thanks 👍 .. now all that i need ..the weekend.. that comes).. to deep dive in sd wan series..

[@habibullahzazai4991]

Hey Rob, Thank you for wonderful series on SDWAN. I had a question, how do we advertise defaul-ruoute (default-originate) from the vEdge to the service VPN, while using template?

[@danilotorlai3916]

On SW16 the routes learned via BGP and next hop is vEdge1 are not working. Because vEdge 1 is choosing the branches' routes via OSPF (110) instead of OMP (250). vEdege 1 is not advertising the OMP routes into OSPF. I don't know if this behaviour is expected.

[@marshallshelton6262]

I think I had some weirdness going on. On vEdge1 with just ospf running and redistributing omp my switch16 was receiving all the correct routes (site3, 4, etc) In the vEdge1 when I enabled BGP with the "redistribute ospf" flag, the original omp routes were not included, seems it didn't want to redistribute routes that were ospf external routes. Redistribution of omp int BGP worked fine. I am also wondering why you would with BGP redistribute the omp routes via their redistribution into ospf (omp->ospf->bgp) seems to be cleaner to just redistribute omp into ospf and just use ospf for loopback and intra area (or site) communications.

[@GlenNetworking]

Yeah not gonna lie, this lab tripped me up big time! But in the end was good learning :)

[@jozefhamar5194]

Not sure if it's just me missing something along the path, but by the end of this video vEdge1's and vEdge2's redistribution is not quite the same, which creates sort of mess. vEdge1: OSPF -> BGP vEdge2: OSPF -> BGP, OMP -> BGP, OMP -> OSPF

[@ghulamrasoulmustafavy7269]

Next hop should be changed because when a Router advertise a route to another iBGP it doesn't change the next-hop and here SW16 doesn't have route to next hop of Edge1 that's why doesn't learn route till change next-hop to self on Edge1 Thank you, you are great🙏

[@joshuawang6953]

Hi Rob, I need your help on this, becasue I am stucking to understand few things. In your Lab 14, I saw you redistributed OMP to OSPF on vEdge2, and vEdge1 SW16 vEdge2 are in the same OSPF area, what will end up is 1. vEdge1 will learn 13.13.13.13/32(example) from OMP, and the same time from OSPF, what going to happen then? 2. Since vEdge1 learned 13.13.13.13/32 via OSPF, and OMP will propagate OSPF by deafut, but it also learned 13.13.13.13/32 via OMP, is this a spilt-brain? In your BGP lab 17, I saw you "show ip route" on SW16, there are only OE2 deafult routes, where are the OE2 routes learned from vEdge2(becasue you just redistributed OMP to OSPF on vEdge2)? Please kindly help me to understand this, I cannot sleep with these questions.

[@RobRikerTechChannel]

Ok, so, you'll need a CCNP or better understanding of OSPF first to understand how OSPF works and then things will click way easier for you. The reason why the O E2 routes show up on SW16 is because the vEdges are using OSPF default information originate. That's just for internet reachability in the lab. IOS13's loopback will be learned from OMP and then be redistributed into OSPF and learned on SW16 via OSPF. I highly suggest you learn OSPF routing first, SD-WAN will be tricky to grasp until you do.

[@joshuawang6953]

@@RobRikerTechChannel Thanks Rob, maybe I didn't make myself clear. I understand how OSPF works, so far the OSPF you are using is not fancy feature, you just propagate a deafult route and redistribute OMP to OSPF. My question is more related to OMP/OSPF route selection and how OMP loop prevent, say vEdge1 will got 13.13.13.13/32 from OSPF OE2, the same time it will got from OMP route from vEdge3, why vEdge1 installed 13.13.13.13/32 from OSPF routes? Second question is since OMP redistribute connected/static/OSPF intra/OSPF inter by deafult, does vEdge1 will follow the rule propagate 13.13.13.13/32(learnt via OSPF) via OMP?

[@joshuawang6953]

@@RobRikerTechChannel I think I figure this out. Two key point, OSPF OE2 routes has distance 110, OMP got 250, so OSPF OE2 got installed to routing table(F,S). 2. vEdge1 will not propagate the route(OSPF OE2) to vSmart, because by default, OMP only automatically send conencted/static/OSPF inter/OSPF intra to vSmart. @Rob, did I get the point there? I would like say it again, I am not talking about OSPF default route or redistribution, I am talking about OMP best route selection and how OMP will propagate routes to vSmart, you didn't cover that in detail in this LAB.

[@RobRikerTechChannel]

@@joshuawang6953 wasn't the point of the lab. You did what every engineer should, find answers to your questions through research. The only way your gonna go anywhere in IT. Now you can get some sleep!

[@joshuawang6953]

@@RobRikerTechChannel Thanks Rob, your series is truely treasure I found in KZbin, although I need spend much time on LAB, but really really enjoy it and not planning to give it up. Now I am in LAB 17, got OSPF part cleared, got every single OSPF routes on all vEdges understood, I think I am ready go to BGP, I will only raise my questions only if I have to, agian thank you Rob.

[@ABUNDANCEandBEYONDATHLETE]

Thanks Rob! 😁

[@你爹-y3f]

hi if i dont any dna devices or some charged net lab accounts ， where can i learn dna by building my own lab just like using eve platform for traditional network lab

[@RobRikerTechChannel]

for DNA you can't on EVE. You'll need dCloud.

[@你爹-y3f]

Rob Riker's Tech Channel yea，thanks ， i just logged in dcloud， i didn't see any thing about dna ， maybe it’s because i dont have enough access level to view it ，～

[@SoulJah876]

Maybe try Cisco devnet sandboxes

[@你爹-y3f]

@@SoulJah876 does it help ？ i have realized that sad lab which in dcould totally won’t help and useless