BGP Local Preference Configuration on Cisco ACI

Topology

Initially, we have Leaf-101 and Leaf-102 having BGP neighborship established with router in ASN 65033. And this router (ASN 65033) is advertising the route 172.16.20.0/24 to both neighbors (Leaf-101 and Leaf-102) which are in ASN 65001.



Initial setup

Verify BGP and routing table on both leafs, we see that the eBGP route is preferred in each leaf and installed in RIB:

Leaf-101:

Leaf-101 has 2 BGP paths for the route “172.16.20.0/24”:

  • one is eBGP via neighbor 192.168.50.2
  • Other is via iBGP, via the fabric and pointing to Leaf-102 TEP 10.0.168.69

The eBGP route is preferred.

leaf-101#  show bgp ipv4 unicast vrf bameur_MC:VRF-MC3
BGP routing table information for VRF bameur_MC:VRF-MC3, address family IPv4 Unicast
BGP table version is 34, local router ID is 1.1.1.1
Status: s-suppressed, x-deleted, S-stale, d-dampened, h-history, *-valid, >-best
Path type: i-internal, e-external, c-confed, l-local, a-aggregate, r-redist, I-injected
Origin codes: i - IGP, e - EGP, ? - incomplete, | - multipath, & - backup

   Network            Next Hop            Metric     LocPrf     Weight Path
* i172.16.20.0/24     10.0.168.69                       100          0 65033 i
*>e                   192.168.50.2                                   0 65033 i

leaf-101# acidiag fnvread | grep 10.0.168.69
     102        1    leaf-102      FDXXXXXX    10.0.168.69/32    leaf         active   0

leaf-101# show ip route vrf bameur_MC:VRF-MC3
IP Route Table for VRF "bameur_MC:VRF-MC3"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

172.16.20.0/24, ubest/mbest: 1/0
    *via 192.168.50.2%bameur_MC:VRF-MC3, [20/0], 00:21:31, bgp-65001, external, tag 65033


Leaf-102:

Same for Leaf-102, it has 2 BGP paths for the route “172.16.20.0/24”:

  • one is eBGP via neighbor 192.168.60.2
  • Other is via iBGP, via the ACI fabric and pointing to Leaf-101 TEP 10.0.240.32
leaf-102# show bgp ipv4 unicast vrf bameur_MC:VRF-MC3
BGP routing table information for VRF bameur_MC:VRF-MC3, address family IPv4 Unicast
BGP table version is 33, local router ID is 2.2.2.2
Status: s-suppressed, x-deleted, S-stale, d-dampened, h-history, *-valid, >-best
Path type: i-internal, e-external, c-confed, l-local, a-aggregate, r-redist, I-injected
Origin codes: i - IGP, e - EGP, ? - incomplete, | - multipath, & - backup

   Network            Next Hop            Metric     LocPrf     Weight Path
* i172.16.20.0/24     10.0.240.32                       100          0 65033 i
*>e                   192.168.60.2                                   0 65033 i

 leaf-102# acidiag fnvread | grep 10.0.240.32
     101        1    leaf-101      FDXXXXXXS     10.0.240.32/32    leaf         active   0

leaf-102# show ip route vrf bameur_MC:VRF-MC3
IP Route Table for VRF "bameur_MC:VRF-MC3"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

172.16.20.0/24, ubest/mbest: 1/0
    *via 192.168.60.2%bameur_MC:VRF-MC3, [20/0], 00:24:03, bgp-65001, external, tag 65033


Leaf-103 (compute leaf): 

leaf-103# show ip route vrf bameur_MC:VRF-MC3
IP Route Table for VRF "bameur_MC:VRF-MC3"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

172.16.20.0/24, ubest/mbest: 2/0
    *via 10.0.168.69%overlay-1, [200/0], 00:00:41, bgp-65001, internal, tag 65033
    *via 10.0.240.32%overlay-1, [200/0], 00:00:41, bgp-65001, internal, tag 65033

leaf-103# show bgp ipv4 unicast vrf bameur_MC:VRF-MC3
BGP routing table information for VRF bameur_MC:VRF-MC3, address family IPv4 Unicast
BGP table version is 102, local router ID is 172.16.30.254
Status: s-suppressed, x-deleted, S-stale, d-dampened, h-history, *-valid, >-best
Path type: i-internal, e-external, c-confed, l-local, a-aggregate, r-redist, I-injected
Origin codes: i - IGP, e - EGP, ? - incomplete, | - multipath, & - backup

   Network            Next Hop            Metric     LocPrf     Weight Path
*|i172.16.20.0/24     10.0.240.32                       100          0 65033 i
*>i                   10.0.168.69                       100          0 65033 i


Note: iBGP/eBGP path and administrative distance

When two BGP paths are compared, the BGP attributes (best path selection process) will determine the best path.
Preference between eBGP vs iBGP path isn’t directly done according to the the administrative
distance (20 or 200), it will always go through the BGP path selection algorithm.
If there is a tie before the “preference of eBGP over iBGP” step, then eBGP path will be selected

Please refer to the following article about BGP best path selection process:

Explore The BGP Path Selection Attributes [Explained with Labs]


Configure BGP Local Preference in ACI

The goal is to prefer Leaf-102 for the route 172.16.20.0/24, to do this, we will modify the local preference value when importing the route from the neighbor of leaf-102, we will set it to a value greater than default (100).

Local preference when importing the route from 192.168.60.2 will be set to 200.

Configuration in ACI:

First, I have create a match and set rules:

1- Match rule: to match for the subnet “172.16.20.0/24”


2- Set Rule: In the set rule, we specify the local preference value as 200:


3- Route-map:

Now, we will create a route map for route control:

Navigate to tenant > Policies > Protocol > Route Maps for Route control:


In the route-map, we select the match route and set rule already created:


It will look like this, and submit:


4- Apply the route map:

Last step is to apply the route-map in the Import direction under the BGP neighbor for leaf-102, that way when receiving the route via the neighbor on leaf-102, it will set the local preference as 200:



Verifications:

  • After applying the route-map in the inbound (import) direction for the Leaf-102 neighbor, we see this the local preference value has changed to 200 for the route 172.16.20.0/24 received from this neighbor (192.168.60.2, connected to Leaf-102).
  • This local preference is propagated inside the ASN 65001 (which is ACI fabric via the Spines which are the Route Reflector). As a result, we see that Leaf-101 is preferring the path via Leaf-102 (pointing to PTEP of leaf-102) instead of directly connected eBGP route.

    So basically, when reaching to this destination (route 172.16.20.0/24), ACI will always exit via leaf-102. In case leaf-102 is down, we


Note

 BGP Route reflector (Spine in case of ACI) will only advertise the best path, that’s why we don’t see path via Leaf-101 being advertised via RR to the fabric.


Leaf-102:

leaf-102# show bgp ipv4 unicast vrf bameur_MC:VRF-MC3
BGP routing table information for VRF bameur_MC:VRF-MC3, address family IPv4 Unicast
BGP table version is 37, local router ID is 2.2.2.2
Status: s-suppressed, x-deleted, S-stale, d-dampened, h-history, *-valid, >-best
Path type: i-internal, e-external, c-confed, l-local, a-aggregate, r-redist, I-injected
Origin codes: i - IGP, e - EGP, ? - incomplete, | - multipath, & - backup

   Network            Next Hop            Metric     LocPrf     Weight Path
*>e172.16.20.0/24     192.168.60.2                      200          0 65033 i

leaf-102# show ip route vrf bameur_MC:VRF-MC3
IP Route Table for VRF "bameur_MC:VRF-MC3"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

172.16.20.0/24, ubest/mbest: 1/0
    *via 192.168.60.2%bameur_MC:VRF-MC3, [20/0], 00:03:59, bgp-65001, external, tag 65033


Leaf-101:

leaf-101#  show bgp ipv4 unicast vrf bameur_MC:VRF-MC3
BGP routing table information for VRF bameur_MC:VRF-MC3, address family IPv4 Unicast
BGP table version is 36, local router ID is 1.1.1.1
Status: s-suppressed, x-deleted, S-stale, d-dampened, h-history, *-valid, >-best
Path type: i-internal, e-external, c-confed, l-local, a-aggregate, r-redist, I-injected
Origin codes: i - IGP, e - EGP, ? - incomplete, | - multipath, & - backup

   Network            Next Hop            Metric     LocPrf     Weight Path
*>i172.16.20.0/24     10.0.168.69                       200          0 65033 i
* e                   192.168.50.2                                   0 65033 i


leaf-101# show ip route vrf bameur_MC:VRF-MC3
IP Route Table for VRF "bameur_MC:VRF-MC3"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

172.16.20.0/24, ubest/mbest: 1/0
    *via 10.0.168.69%overlay-1, [200/0], 00:04:12, bgp-65001, internal, tag 65033


Leaf-103 (compute leaf): 

leaf-103# show ip route vrf bameur_MC:VRF-MC3
IP Route Table for VRF "bameur_MC:VRF-MC3"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

172.16.20.0/24, ubest/mbest: 1/0
    *via 10.0.168.69%overlay-1, [200/0], 00:46:15, bgp-65001, internal, tag 65033

leaf-103# show bgp ipv4 unicast vrf bameur_MC:VRF-MC3
BGP routing table information for VRF bameur_MC:VRF-MC3, address family IPv4 Unicast
BGP table version is 98, local router ID is 172.16.30.254
Status: s-suppressed, x-deleted, S-stale, d-dampened, h-history, *-valid, >-best
Path type: i-internal, e-external, c-confed, l-local, a-aggregate, r-redist, I-injected
Origin codes: i - IGP, e - EGP, ? - incomplete, | - multipath, & - backup

   Network            Next Hop            Metric     LocPrf     Weight Path
*>i172.16.20.0/24     10.0.168.69                       200          0 65033

Bilel-A

Enthusiastic Network Engineer specializing in Cisco ACI, passionate about solving challenges. A lifelong learner who loves gaining and sharing knowledge. Profile: https://www.linkedin.com/in/bilel-ameur-71116b2b5

Related articles

MACsec Explained and Configuration Example

BGP States Troubleshooting – BGP Basics

VXLAN EVPN L2 VNI, L3VNI Configuration and Troublshooting

VRF Route Leaking on Cisco Nexus Switches [Explained & Configuration]

BFD – Bidirectional Forwarding Detection Explained & configuration

OTV Failure isolation

EPLD Upgrade on Nexus and MDS switches

Designated Forwarder (DF) and PIM Assert Mechanism Explained

Understanding Multicast PIM Sparse Mode (Packet Flow)

Understanding MST: Multiple Spanning Tree Protocol With Example [GNS3 Lab]

0 0 votes
Article Rating
Subscribe
Notify of
guest
0 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
0
Would love your thoughts, please comment.x
()
x