AW: [nsp] Size of BGP Table from ISP-X

Ryan O'Connell ryan-nsp at complicity.co.uk
Sun Feb 23 18:55:39 EST 2003


On Sun, 23 Feb 2003 19:21:43 +0100 Peter Burggasser <p.burggasser at uta1002.at> wrote:
> If the table needs more memory as my router have. Are there mechanism to
> reduce the table? 

You can just receive default routes from your upstreams or you can just
drop all prefixes except half a dozen or so and point the default route at
them. (Getting default only from your upstream won't help you if there's a
break in their core or their BGP setup)

If you want something other than an active/backup failover set up, you can
take just customer routes from a number of ISPs - so traffic will take a
direct path if one is available - but not take all routes. Then use one of
the methods above for the default route.

If your router can take it, it's best to have all routes and then some
memory spare - this is so you can use soft-reconfiguration inbound and
change routing policies without bouncung BGP sessions.

-- 
      Ryan O'Connell
Mail: ryan at complicity.co.uk
  CV: http://www.complicity.co.uk/ryancv.pdf
      CCIE #8174



More information about the cisco-nsp mailing list