CCIE - Internetwork Expert's Online Community

hi guys, this is my first post with my first question:

Basically, can someone explain to me the bellow `sh ip bgp sum`s ?

The setup is quite simple: 2x 2811 routers that connect to 2x ISPs and that have themselves (the routers) set as iBGP Peers. Filters are only set on the out to isp (so that i don't become a transit AS), not between the iBGP peers. HSRP enabled on both internal int of the routers (BGP shouldn't care though about it)

Question1: x.x.x.76 gets 268025 prefixes (subnets) from neighbor r.r.r.r & x.x.x.96 gets 268323 prefixes (subnets) from s.s.s.s. Why does x.x.x.96 say that he gets only 119006 prefixes (routes) from x.x.x.76 and not the 268025 that x.x.x.76 from r.r.r.r ? Same for the other Router (x.x.x.96).

Question 2: shouldn't the maximum of bgp activity paths equal (example for x.x.x.96) to 268323+119006+1(=287330) and not the 4391402 that it reports ?

Question 3: A prefix is a subnet in the `sh ip bgp` table right ?

Question 4: A path is every row in the `sh ip bgp` table for a distinct prefix/subnet, right ?

Question 5: What are (also example for x.x.x.96) 316682/46726 & 4778733/4391402 ?

Question 6(most important): If these two routers are connected to 2 ISPs via 2 different bandwidth connections (let's say x.x.x.96 has more bandwith to the metropolitan network), how can i tweak bgp so that all traffic (incoming&outgoing) from/to metro IPs comes/goes via him and not x.x.x.76 ?

On x.x.x.76 i get:

BGP activity 280508/10567 prefixes, 887554/386303 paths, scan interval 60 secs

Neighbor        V    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down  State/PfxRcd
x.x.x.96  4  <AS>  102061   89429  1454987    0    0 17:00:10   233225
r.r.r.r  4  <AS>  102290      57  1454938    0    0 00:28:27   268025

on x.x.x.96 i get:

BGP activity 316682/46726 prefixes, 4778733/4391402 paths, scan interval 60 secs

Neighbor        V    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down  State/PfxRcd
y.y.y.y     4 <AS>    1683   69378  7085866    0    0 20:14:22        1
x.x.x.76    4 <AS>  802772  726567  7085920    0    0 16:58:53   119006
s.s.s.s  4 <AS>  238859   69264  7085866    0    0 20:14:20   268323

1. Because .76 has received routes from .96 that are a better choice than
its link to the ISP. Much like any routing protocol, if you learn from
NeighborB the best path, there is no sense in readvertising what NeighborA
sends you about the same route!

You'll advertise best paths.

2. 268323 + 119006 + 1 is 387330. But your prefixes may overlap. You can
learn the identical prefix from two places, you'll then pick the best path
to install/readvertise. So your total learned path may vary from time to
time!

3. Yup. Unique advertisement.

4. Pretty much. Although there are summary lines, and obvious duplicates
as well.

5. This has to do with memory allocation for a path. How many times have
there been adds/withdrawals/updates in the time the peers have been up.

6. BGP, by default, does not take bandwidth into account. There are new
features regarding "dmzlink" that can help take bandwidth into account.

HTH,


Scott Morris, CCIE4 #4713, JNCIE-M #153, JNCIS-ER, CISSP, et al.
CCSI/JNCI-M/JNCI-ER
Senior CCIE Instructor

smorris@internetworkexpert.com



Internetwork Expert, Inc.
http://www.InternetworkExpert.com
Toll Free: 877-224-8987
Outside US: 775-826-4344
Online Community: http://tinyurl.com/6dmnsu
CCIE Blog: http://tinyurl.com/2nxxaq

Knowledge is power.
Power corrupts.
Study hard and be Eeeeviiiil......


________________________________

From: ccie-rs@ieoc.com [mailto:ccie-rs@ieoc.com] On Behalf Of ardealul
Sent: Thursday, November 20, 2008 5:59 AM
To: smorris@internetworkexpert.com
Subject: [CCIE R&S] bgp "newbie" question(s)



hi guys, this is my first post with my first question:

Basically, can someone explain to me the bellow `sh ip bgp sum`s ?

The setup is quite simple: 2x 2811 routers that connect to 2x ISPs and that
have themselves (the routers) set as iBGP Peers. Filters are only set on the
out to isp (so that i don't become a transit AS), not between the iBGP
peers. HSRP enabled on both internal int of the routers (BGP shouldn't care
though about it)

Question1: x.x.x.76 gets 268025 prefixes (subnets) from neighbor r.r.r.r &
x.x.x.96 gets 268323 prefixes (subnets) from s.s.s.s. Why does x.x.x.96 say
that he gets only 119006 prefixes (routes) from x.x.x.76 and not the 268025
that x.x.x.76 from r.r.r.r ? Same for the other Router (x.x.x.96).

Question 2: shouldn't the maximum of bgp activity paths equal (example for
x.x.x.96) to 268323+119006+1(=287330) and not the 4391402 that it reports ?

Question 3: A prefix is a subnet in the `sh ip bgp` table right ?

Question 4: A path is every row in the `sh ip bgp` table for a distinct
prefix/subnet, right ?

Question 5: What are (also example for x.x.x.96) 316682/46726 &
4778733/4391402 ?

Question 6(most important): If these two routers are connected to 2 ISPs via
2 different bandwidth connections (let's say x.x.x.96 has more bandwith to
the metropolitan network), how can i tweak bgp so that all traffic
(incoming&outgoing) from/to metro IPs comes/goes via him and not x.x.x.76 ?

On x.x.x.76 i get:

BGP activity 280508/10567 prefixes, 887554/386303 paths, scan interval 60
secs

Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down
State/PfxRcd
x.x.x.96 4 102061 89429 1454987 0 0 17:00:10 233225
r.r.r.r 4 102290 57 1454938 0 0 00:28:27 268025

on x.x.x.96 i get:

BGP activity 316682/46726 prefixes, 4778733/4391402 paths, scan interval 60
secs

Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down
State/PfxRcd
y.y.y.y 4 1683 69378 7085866 0 0 20:14:22 1
x.x.x.76 4 802772 726567 7085920 0 0 16:58:53 119006
s.s.s.s 4 238859 69264 7085866 0 0 20:14:20 268323




Internetwork Expert - The Industry Leader in CCIE Preparation
http://www.internetworkexpert.com

Subscription information may be found at:
http://www.ieoc.com/forums/ForumSubscriptions.aspx