[isp-bgp] Re: Smaller ISP, do we need full BGP routes?

[David Raistrick <drais@...]

> shouldn't matter. We should just take partial or
> default routes and let our backbone providers decide
> where to send the outbound traffic. Our inbound
> traffic is determined through our AS prepending &
> advertisements config.


I just switched from two full views down to two defaults.  (the smallest
latest-bug-fixed IOS for my 1760 used about 10 megs more memory then then
old..)

I just filtered the recieved routes, leaving only the defaults. (I had
both providers sending full tables + defaults.)

Using the BGP-received defaults all of my outbound traffic prefers one
link:

   Network          Next Hop            Metric LocPrf Weight Path
r  0.0.0.0          65.112.128.1                           0 209 i
r>                  144.232.153.161          1             0 1239 i



The "prefered" route seems to vary depending on which provider's table
comes in first.

I haven't taken the time to look, but there may be a way to force the
metrics to be even.  I dunno.

In the meantime I installed static defaults:

S*   0.0.0.0/0 [1/0] via 144.232.153.161
               [1/0] via 65.112.128.1

This gives me the connection-based balancing I'm looking for.

FWIW, we're NOT an ISP...most of our traffic is outbound.

A snippit from my notes (sorry, i don't recall who to attribute this to):

"Per-destination or per-packet load-balancing depends on the type of
switching scheme used for IP packets. By default, on most Cisco routers,
fast switching is enabled under interfaces. This is a demand caching
scheme that does per-destination load-balancing. To set per-packet
load-balancing, enable process switching (or disable fast switching) using
the following commands:

     Router# config t
     Router(config)# interface Ethernet 0
     Router(config-if)# no ip route-cache
     Router(config-if)# ^Z

Now the router's CPU looks at every single packet and load balances on the
number of routes in the routing table for the destination. This
can crash a low-end router because the CPU must do all the processing. To
re-enable fast switching, use these commands:

     Router# config t
     Router(config)# interface Ethernet 0
     Router(config-if)# ip route-cache
     Router(config-if)# ^Z
"



---
david raistrick
drais@atlasta.net		http://www.expita.com/nomime.html


------------------------ANNOUNCEMENT---------------------------------
---------------------------------------------------------------------
           ISPCON FALL 2003 - Santa Clara Convention Center
     October 20-22 - The Definitive Event Wired and Wireless ISPs
                            www.ispcon.com
----------------------------------------------------------------------
----------------------------------------------------------------------