- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-17-2021 03:22 AM
Hi Everyone,
May i know how IBGP causing loop? if all the router is at one AS Number. Thanks !
Solved! Go to Solution.
- Labels:
-
Other Routing
Accepted Solutions
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-17-2021
06:47 AM
- last edited on
12-03-2021
03:20 AM
by
Translator
Hello
For full ibgp mesh bgp uses as stated split horizon rule which won’t advertise ibgp routes to other ibgp peers
for non full mesh ibgp peering then Route-Reflectors are applicable and they utilise attributes such as:
Originator ID = which is the rtr that originated/advertised a prefix
Cluster-id = which is the ID of the cluster list of the RR
If either of these attributes related to the rtr that created them are seen in a prefix on the same rtr then it’s discarded thus negating any loop
Please rate and mark as an accepted solution if you have found any of the information provided useful.
This then could assist others on these forums to find a valuable answer and broadens the community’s global network.
Kind Regards
Paul
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-17-2021 03:43 AM - edited 09-17-2021 03:46 AM
Hi,
Have you been diagnosed with a loop in your network topology ?
If so, I would appreciate it if you could expand a little on the topology, and how you came to the conclusion that there is indeed a Loop.
If not, and you want to know how iBGP avoided loop then:
BGP prevents loops in two ways:
Do not accept routes with your own AS from eBGP peers.
Horizon rules for iBGP Peers:
a)Do announce eBGP routes to all peers.
b)Do announce iBGP routes to external peers.
c)Do not announce iBGP routes to internal peers.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-17-2021 03:54 AM
This only happends when you have full mesh connectivity, IGP to BGP and BGP to IGP restribution take place.
which can avoice using Route-map to stop doing that.
Note : again depends on the network design.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-17-2021 06:10 AM
Basically im watching a ccnp course from udemy, which is chinese one. he didnt exactly talk about how ibgp causing the loop. Just saying use full mesh to prevent the loop
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-17-2021 02:13 PM
I think the instructor meant to say that because of loop=prevarication mechanism of BGP (listed above), you must use full mesh topology (or route-reflector) to have proper connectivity and reachability.
Regards, ML
**Please Rate All Helpful Responses **
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-17-2021
06:47 AM
- last edited on
12-03-2021
03:20 AM
by
Translator
Hello
For full ibgp mesh bgp uses as stated split horizon rule which won’t advertise ibgp routes to other ibgp peers
for non full mesh ibgp peering then Route-Reflectors are applicable and they utilise attributes such as:
Originator ID = which is the rtr that originated/advertised a prefix
Cluster-id = which is the ID of the cluster list of the RR
If either of these attributes related to the rtr that created them are seen in a prefix on the same rtr then it’s discarded thus negating any loop
Please rate and mark as an accepted solution if you have found any of the information provided useful.
This then could assist others on these forums to find a valuable answer and broadens the community’s global network.
Kind Regards
Paul
