From: Victor Cappuccio (cvictor@protokolgroup.com)
Date: Wed Jun 14 2006 - 03:48:07 ART
Petr thanks for your reply, By the Telephone Book: the Multi Exit
Discriminator (MED) value on prefixes advertised to external BGP (eBGP)
neighbors to match the Interior Gateway Protocol (IGP) metric of the next
hop, we must use the set metric-type internal command in route-map and to
return to the default, we use the no form of this command, so there is no
default metric for route advertised.
My Question is why attributes like the Metric are not propagated for some
routes but for others
Look at this output
*Mar 1 13:41:36.790: BGP(0): 145.1.136.1 rcvd UPDATE w/ attr: nexthop
145.1.136.1, origin i, localpref 100
And the difference with this output
*Mar 1 13:41:36.790: BGP(0): 145.1.136.1 rcvd UPDATE w/ attr: nexthop
145.1.136.1, origin i, localpref 100
Im looking for some command on the remote-side, to send the metric value of
0 (this seems to be contradictory by my interpretation with RFC 4451
<http://www.ietf.org/rfc/rfc4451.txt> http://www.ietf.org/rfc/rfc4451.txt
For example, earlier versions of the specification called for a missing MED
to be assigned the highest possible MED value (i.e., 2^32-1). ) .. Yes we
have bgp bestpath med missing-as-worst command but Im starting to think
that confusion is part this journey.
Thanks Again
Victor.
_____
De: Petr Lapukhov [mailto:petrsoft@gmail.com]
Enviado el: Miircoles, 14 de Junio de 2006 02:19 a.m.
Para: Victor Cappuccio
CC: ccielab@groupstudy.com
Asunto: Re: Set BGP Metic in Routes
Victor,
Usually, when you originate a prefix into BGP, it make metric attribute
from IGP metric. It does not matter if this is hop count or OSPF cost,
or EIGRP composite metric :)
When you do some "internal" prefix generation, like default-originate,
you may not have metric attribute at all :) It's optional, after all :)
HTH
Petr
2006/6/14, Victor Cappuccio <cvictor@protokolgroup.com>:
Hi Guys,
Why sometimes when checking the BGP Table we can observe routes with a
metric value set and sometimes not
Like in this case:
BGP routing table entry for 0.0.0.0/0, version 7
Paths: (2 available, best #1, table Default-IP-Routing-Table)
Multipath: iBGP
Advertised to non peer-group peers:
145.1.37.7 145.1.53.5 145.1.136.6
Local, (Received from a RR-client)
145.1.136.1 from 145.1.136.1 (150.1.1.1) <<<<< === See there is
no metric value set
Origin IGP, localpref 100, valid, internal, multipath, best
Local, (Received from a RR-client)
145.1.136.6 from 145.1.136.6 (150.1.6.6)
Origin IGP, metric 0, localpref 100, valid, internal, multipath <<<<<
=== the metric is set to 0
R3 is a RR peering with R1 and R6 in the same Ethernet segment
And the configuration in R1 and R6 looks like this:
Rack1R6#show run | be router bgp
router bgp 100
bgp router-id 150.1.6.6
bgp log-neighbor-changes
neighbor 54.1.10.254 <http://54.1.10.254> remote-as 54
neighbor 145.1.136.3 remote-as 100
!
address-family ipv4
neighbor 54.1.10.254 activate
neighbor 54.1.10.254 dmzlink-bw
neighbor 145.1.136.3 activate
neighbor 145.1.136.3 send-community both
neighbor 145.1.136.3 next-hop-self
neighbor 145.1.136.3 default-originate
neighbor 145.1.136.3 route-map DENYALL out
no auto-summary
no synchronization
exit-address-family
Rack1R1#show run | b router bgp
router bgp 100
bgp router-id 150.1.1.1
bgp log-neighbor-changes
neighbor 145.1.136.3 remote-as 100
neighbor 204.12.1.254 remote-as 54
!
address-family ipv4
neighbor 145.1.136.3 activate
neighbor 145.1.136.3 next-hop-self
neighbor 145.1.136.3 send-community both
neighbor 145.1.136.3 default-originate
neighbor 145.1.136.3 route-map DENYALL out
neighbor 204.12.1.254 activate
neighbor 204.12.1.254 dmzlink-bw
no auto-summary
no synchronization
bgp dmzlink-bw
exit-address-family
Rack1R3#show run | be router bgp
router bgp 100
bgp router-id 150.1.3.3
bgp always-compare-med
bgp log-neighbor-changes
neighbor 145.1.37.7 remote-as 100
neighbor 145.1.53.5 remote-as 200
neighbor 145.1.136.1 remote-as 100
neighbor 145.1.136.6 remote-as 100
maximum-paths ibgp 2
!
address-family ipv4
neighbor 145.1.37.7 activate
neighbor 145.1.37.7 route-map ONLYDEFA out
neighbor 145.1.53.5 activate
neighbor 145.1.53.5 route-map ONLYDEFA out
neighbor 145.1.136.1 activate
neighbor 145.1.136.1 route-reflector-client
neighbor 145.1.136.1 send-community both
neighbor 145.1.136.6 activate
neighbor 145.1.136.6 route-reflector-client
neighbor 145.1.136.6 send-community both
maximum-paths ibgp 2
no auto-summary
no synchronization
bgp dmzlink-bw
network 145.1.136.0 mask 255.255.255.0
aggregate-address 145.1.0.0 255.255.0.0 <http://255.255.0.0> summary-only
exit-address-family
Any opinion here please
Thanks
Victor.
This archive was generated by hypermail 2.1.4 : Sat Jul 01 2006 - 07:57:32 ART