From: Stefan L. Dozier (doziersl@yahoo.com)
Date: Thu Oct 31 2002 - 22:24:07 GMT-3
Pete, thanks for the reply!
And I would have the tendency to agree, that's it's probably not
that abnormal, considering the behavior of what the lab was trying
to teach, worked as expected. But it just piqued my interest when
the bgp table wasn't what I expected to see, and since I couldn't
find any definitive info explaining why, figured I maybe overlooked
something stupid!
Anyway, since you mentioned the bgp rib, I have did a bit more research
and came up with this excerpt.....
"BGP installs up to the maximum number of paths allowed (configured using
the maximum-paths command). BGP uses the best path algorithm to select one
multipath as the best path, insert the best path into the routing
information base (RIB), and advertise the best path to BGP peers. Other
multipaths may be inserted into the RIB, but only one path will be selected
as the best path".
Here's the link that provides this info, with the caveat that it's
in reference to "BGP Multipath Load Sharing for Both eBGP and iBGP
in an MPLS VPN" using 12.2(4)T code and I'm using 12.1(5)T9 and not
doing any MPLS!
Watch the word wrap.....
http://www.cisco.com/en/US/products/sw/iosswrel/ps1839/products_feature_guid
e09186a0080087c21.html
Stefan
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
Peter van Oene
Sent: Thursday, October 31, 2002 8:21 AM
To: Groupstudy CCIELab List
Subject: Re: "BGP Maximum-Paths"
Since it is using both, it might simply be marking the first in the
list. I'd have to mess with it a bit to figure out if it actually marks
two paths in the bgp rib, but what you are seeing doesn' t seem that
abnormal.
Pete
At 09:15 PM 10/30/2002 -0500, Stefan L. Dozier wrote:
>I'm working from Parkhurst's BGP-4 Command and Configuration
>Handbook, chapter 7 pages 114-117 with Figure 7-1 labbed up!
>
>I understand the "best-path algorithm" and that by default BGP
>only installs the best path in the ip routing table. I also
>understand the value of the maximum-paths command in that it
>allows up to six paths to the same dest to be installed in the
>ip routing table.
>
>On to the lab scenario!
>
>Initially config'd, the bgp table and ip routing table on
>RouterA displays expects results.
>
>RouterA#sh ip bgp
>BGP table version is 20, local router ID is 10.1.2.1
>Status codes: s suppressed, d damped, h history, * valid, > best, i -
>internal
>Origin codes: i - IGP, e - EGP, ? - incomplete
>
> Network Next Hop Metric LocPrf Weight Path
>* 172.17.1.0/24 10.1.2.2 0 0 1 i
>*> 10.1.1.2 0 0 1 i
>
>RouterA#sh ip route bgp
> 172.17.0.0/24 is subnetted, 1 subnets
>B 172.17.1.0 [20/0] via 10.1.1.2, 00:10:00
>
>After the "maximum-paths 2" command was added to the bgp process
>on RouterA.
>
>RouterA#sh ip bgp
>BGP table version is 21, local router ID is 10.1.2.1
>Status codes: s suppressed, d damped, h history, * valid, > best, i -
>internal
>Origin codes: i - IGP, e - EGP, ? - incomplete
>
> Network Next Hop Metric LocPrf Weight Path
>*> 172.17.1.0/24 10.1.2.2 0 0 1 i
>* 10.1.1.2 0 0 1 i
>RouterA#sh ip route bgp
> 172.17.0.0/24 is subnetted, 1 subnets
>B 172.17.1.0 [20/0] via 10.1.2.2, 00:01:38
> [20/0] via 10.1.1.2, 00:01:38
>
>While the ip routing table displays expected results, I can't
>find any information on why when I use the maximum-paths command
>it seems to break the best-path algorithm and chooses the path
>with the higher router-id as indicated above! And since Parkhurst
>didn't find it necessary to paste the bgp table after insertion
>of the maximum-paths command under the BGP process, I'm kinda
>suspicious! I am missing something here?
>
>Stefan
This archive was generated by hypermail 2.1.4 : Tue Nov 05 2002 - 08:36:01 GMT-3