12-27-2023 10:02 AM
Solved! Go to Solution.
12-27-2023 09:43 PM
Exactly,
there is no such step in OMP bestpath selection:
12-27-2023 10:05 AM
Can I see topolgy
MHM
12-27-2023 10:09 AM
lab-r1 has SDWAN tunnels to both irv-r1 and chr-r1, which learned the route from BGP on the LAN side. Is this what you are asking for?
12-27-2023 11:06 AM - edited 12-27-2023 02:17 PM
MHM
12-27-2023 11:24 AM
It's not the same. The issue is OMP is learning that route with different AS length. Why is it showing the same AS length for both?
12-27-2023 02:00 PM
Hi
please share "show bgp vpnv4 unicast vrf 13 2611:6bC0:6000::/40"
12-27-2023 02:07 PM - edited 12-27-2023 02:09 PM
BGP Table on the lab router is showing the correct path. However, the routing table is using OMP routes. See above.
lab-r1#sh bgp vpnv6 unicast vrf 13 2611:6bC0:6000::/40
BGP routing table entry for [1:13]2611:6bC0:6000::/40, version 5832
Paths: (1 available, best #1, table 13)
Not advertised to any peer
Refresh Epoch 1
Local
::FFFF:10.200.34.1 (via vrf 13) from 0.0.0.0 (172.29.236.1)
Origin incomplete, metric 1000, localpref 50, valid, sourced, best
Extended Community: SoO:0:14009
rx pathid: 1, tx pathid: 0x0
Updated on Dec 27 2023 21:15:33 UTC
12-27-2023 10:19 PM
And you have one route in BGP because it is redistribution of the best route.
Even though you have ECMP on OMP side, in BGP you have one route which is "best of the best" using tie-breakers.
But still there is question: on which devices you have enabled "propagate-as" on bgp config? And can you confirm that you have "overlay-as" in OMP configuration?
12-27-2023 02:17 PM
It looks like OMP doesn't take AS-PATH length into route selection process.
12-27-2023 02:25 PM
12-27-2023 09:43 PM
Exactly,
there is no such step in OMP bestpath selection:
12-27-2023 02:28 PM
1. Yes.
2. Yes
lab-r1#sh ipv route vrf 13 2611:6bC0:6000::/40 pro
Routing entry for 2611:6bC0:6000::/40
Known via "omp", distance 251, metric 0, type overlay
Redistributing via bgp 395191
Route count is 2/2, share count 0
Routing paths:
10.200.34.1%default, Sdwan-system-intf%default
From ::FFFF:10.200.34.1
opaque_ptr 0x7F530E63F330
pdb 0x7F5307CE0580, route 0x7F530BC849A8, adj 0x7F530EA92158
OMP attribute 0x7F530E63F330, ref 5
aspath 0x7F530D2D5758, ref 2, length 18, value 13979 65051 65051 65051
Total OMP attr count 7, aspath 7, community 0
Last updated 00:00:50 ago
10.201.34.1%default, Sdwan-system-intf%default
From ::FFFF:10.201.34.1
opaque_ptr 0x7F530E63F330
pdb 0x7F5307CE0580, route 0x7F530BC849A8, adj 0x7F530EA95CA8
OMP attribute 0x7F530E63F330, ref 5
aspath 0x7F530D2D5758, ref 2, length 18, value 13979 65051 65051 65051
Total OMP attr count 7, aspath 7, community 0
Last updated 00:00:50 ago
lab-r1#sh sdwan omp ipv6-routes 2611:6bC0:6000::/40
---------------------------------------------------
omp route entries for vpn 13 route 2611:6bC0:6000::/40
---------------------------------------------------
RECEIVED FROM:
peer 10.131.1.1
path-id 17356
label 1005
status C,I,R
loss-reason not set
lost-to-peer not set
lost-to-path-id not set
Attributes:
originator 10.200.34.1
type installed
tloc 10.200.34.1, private1, ipsec
ultimate-tloc not set
domain-id not set
overlay-id 1
site-id 200
preference not set
tag not set
origin-proto iBGP
origin-metric 0
as-path "13979 65051 65051 65051"
community not set
unknown-attr-len not set
RECEIVED FROM:
peer 10.131.1.1
path-id 17360
label 1003
status C,I,R
loss-reason not set
lost-to-peer not set
lost-to-path-id not set
Attributes:
originator 10.201.34.1
type installed
tloc 10.201.34.1, private1, ipsec
ultimate-tloc not set
domain-id not set
overlay-id 1
site-id 201
preference not set
tag not set
origin-proto iBGP
origin-metric 0
as-path "65000 19852 19852"
community not set
unknown-attr-len not set
12-27-2023 02:36 PM
Can i see config of lab router
MHM
12-27-2023 10:00 PM - edited 12-27-2023 10:22 PM
show omp tlocs [ detail ] [ color lte ] [ encap ipsec ] [ ip ip-address ] [ advertised ] [ received ]
use this to see if there is route for 10.201.34.1 or not
even if the OMP dont select short path we need to see it in BGP
but as I see there is only one path
lab-r1#sh bgp vpnv6 unicast vrf 13 2611:6bC0:6000::/40
BGP routing table entry for [1:13]2611:6bC0:6000::/40, version 5832
Paths: (1 available, best #1, table 13)
Not advertised to any peer
Refresh Epoch 1
Local
::FFFF:10.200.34.1 (via vrf 13) from 0.0.0.0 (172.29.236.1)
Origin incomplete, metric 1000, localpref 50, valid, sourced, best
Extended Community: SoO:0:14009
rx pathid: 1, tx pathid: 0x0
Updated on Dec 27 2023 21:15:33 UTC
12-27-2023 11:12 PM
One path in service side BGP is normal, it is redistributed route. Best is selected (using tie-breakers among ECMP routes) and one is redistributed only.
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