Mailing List Archive: eBGP/iBGP question

We are trying to figure out something with our BGP/OSPF setup.

We have two routers - each with a connection abroad hearing the full
800K prefix table via eBGP.
These two routers are connected via iBGP (as well as OSPF), yet
router1 only hears 19K prefixes from
router0 whereas router0 hears all 800K prefixes from router1:

Router0:
Neighbor Spk AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down
St/PfxRcd
62.40.125.1 0 20965 5793423 1835150 454926506 0 0 5w4d
18791
83.97.88.89 0 21320 109164804 1835118 454926506 0 0
5w4d 813154
128.139.220.90 0 378 102799517 102295152 454926506 0 0
15w6d 804788

Router1:
Neighbor Spk AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down
St/PfxRcd
62.40.125.65 0 20965 1879521 662142 174748912 0 0 05:38:56
18792
83.97.88.93 0 21320 38692259 568243 174748912 0 0
05:39:04 813161
128.139.220.91 0 378 29861127 28724590 174748912 0 0
15w6d 18799

How can we get Router1 which peers with 128.139.220.91 to hear the
full routing table?

Ideas and clues welcome.

Relevant configs below.

Thanks,
Hank

Router0 config:

router bgp 378
!
bgp log neighbor changes detail
address-family ipv4 unicast
bgp dampening 5 750 3000 10
redistribute static route-policy static-to-bgp
redistribute ospf 378 metric 254 route-policy BezeqINT_LNS_BGP
!
route-policy static-to-bgp
! Sets BGP community
apply BGPv4-blackhole
end-policy
!
route-policy BezeqINT_LNS_BGP
if (destination in pfx_IUCC_LNS_p1_permit) then
set community (378:0)
endif
end-policy
!
prefix-set pfx_IUCC_LNS_p1_permit
10.64.10.2/32
end-set
!
neighbor-group ILAN
remote-as 378
address-family ipv4 unicast
next-hop-self
!
neighbor 128.139.220.90
use neighbor-group ILAN
update-source Loopback0
address-family ipv4 unicast
maximum-prefix 1200000 75
soft-reconfiguration inbound always
!
---------------------------------------------------

Router1

!
router bgp 378
address-family ipv4 unicast
bgp dampening 5 750 3000 10
redistribute static route-policy static-to-bgp
!
neighbor 128.139.220.91
use neighbor-group ILAN
update-source Loopback0
address-family ipv4 unicast
maximum-prefix 1200000 75
route-policy default_policy_pass_all out
soft-reconfiguration inbound always
!
!
neighbor-group ILAN
remote-as 378
address-family ipv4 unicast
next-hop-self
!
route-policy static-to-bgp
! Just sets community
apply BGPv4-blackhole
end-policy
!
route-policy default_policy_pass_all
pass
end-policy

_______________________________________________
cisco-nsp mailing list cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Hi,

On Thu, Dec 24, 2020 at 10:16:08AM +0200, hank@interall.co.il wrote:
> Ideas and clues welcome.

BGP will only ever signal the best path onward.

So if R1 decides "I like eBGP paths behind R2 much more!", it will never
send its own eBGP paths to R2, because they are not "best".

There is "bgp best-external" and "bgp add-paths" to achieve other
results, but those have implications - like "more memory usage" and
on XR "best-external does totally weird things unless you run labeled-
unicast".

gert
--
"If was one thing all people took for granted, was conviction that if you
feed honest figures into a computer, honest figures come out. Never doubted
it myself till I met a computer with a sense of humor."
Robert A. Heinlein, The Moon is a Harsh Mistress

Gert Doering - Munich, Germany gert@greenie.muc.de

Mailing List Archive

Mailing List Archive

Attached Files: