05-02-2022 11:19 PM
Hi all
In the below design - r1 is doing ebgp peering with r2 and r3
r2 and r3 is advertising 10.10.10.0/24 to r1
r1 -> r3 is the selected path to reach 10.10.10.0/24
Assuming that prefix filtering of any kind is NOT setup
q1) is it normal for r1 to be re-advertising 10.10.10.0/24 to r2 ?
Thank you
Solved! Go to Solution.
05-02-2022 11:46 PM
Perhaps there is something in this question that I am not understanding. But it seems to me that it is pretty straight forward. r3 will advertise 10.10.10.0/24 to r1 as an EBGP route. r1 will advertise that network to r2. r2 will receive the advertisement, will see its own AS number in the advertisement and will not process the advertisement. This is a basic loop avoidance behavior of BGP.
05-03-2022 12:03 AM - edited 05-03-2022 12:04 AM
05-04-2022 10:23 AM - edited 05-04-2022 10:24 AM
1) Yes it isn't R1 because R1 would advertise to R2 in your scenario and it would be R2 who drops it as it is the one seeing it's own AS in the path.
2) All other things being equal (which it should be based on the basic information in your scenario) it would be the one that is received first - see this link for full details -
https://www.cisco.com/c/en/us/support/docs/ip/border-gateway-protocol-bgp/13753-25.html
Jon
05-04-2022 11:43 AM
05-04-2022 11:57 AM - edited 05-04-2022 11:57 AM
05-02-2022 11:46 PM
Perhaps there is something in this question that I am not understanding. But it seems to me that it is pretty straight forward. r3 will advertise 10.10.10.0/24 to r1 as an EBGP route. r1 will advertise that network to r2. r2 will receive the advertisement, will see its own AS number in the advertisement and will not process the advertisement. This is a basic loop avoidance behavior of BGP.
05-02-2022 11:58 PM
R1 will advertise prefix learn from R3 to R2.
R2 will drop this prefix update get from R1 becuase it see our AS in AS-path attribute of perfix.
05-03-2022 12:03 AM - edited 05-03-2022 12:04 AM
Hi,
How EBGP loop prevention work :
05-04-2022
01:09 AM
- last edited on
05-17-2022
05:14 AM
by
Translator
Normally BGP will drop anyhting that comes from it is own AS but there is a command to do this for you we usually use them in DMVPN to let our spokes talk to each other though HUB.
neighbor x.x.x.x allowas-in
or
neighbor x.x.x.x as-override
05-04-2022 10:12 AM
hi all
Appreciate your response
Yeap i do understand that there will be some kind of loop avoidance in place
1) Just wondering which router actually does the loop avoidance ( which i understand from the replies that it isn't R1 )
- R1 receive same routes from 2 different routers with the same AS number and advertised prefix - decide to not re-advertise
- R1 will still re-advertise but either R2 / R3 will drop it
2) what is the tie breaker for R1 path selection if R2 and R3 has the same ASN and advertised the same prefix / path to R1 ?
Thank you
05-04-2022 10:23 AM - edited 05-04-2022 10:24 AM
1) Yes it isn't R1 because R1 would advertise to R2 in your scenario and it would be R2 who drops it as it is the one seeing it's own AS in the path.
2) All other things being equal (which it should be based on the basic information in your scenario) it would be the one that is received first - see this link for full details -
https://www.cisco.com/c/en/us/support/docs/ip/border-gateway-protocol-bgp/13753-25.html
Jon
05-04-2022 11:43 AM
R1
if all criteria tie then
Lowest router-id will tie breaker.
05-04-2022 11:57 AM - edited 05-04-2022 11:57 AM
Yes, if both advertisments were received at the same time.
Jon
05-08-2022 09:28 AM
thank you richard, jon, mhm and all
glad to see that you guys are still around
05-09-2022 09:03 AM
You are welcome. Glad to see you still active in the community.
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