Hi all,
I work at an ISP and we are very interested in running OpenBGPD on the
edges talking to our transport routers. They won't be routing traffic, but
really just act as an internal BGP cache. Right now our Cisco equipment is
not pulling its weight. When we have flaps with an upstream provider we
run into serious downtime... It takes Cisco about 10 minutes to crunch the
bgp tables whereas our tests so far show that OpenBGPD does it in about 8
seconds. This is fantastic!
We have a test environment setup but there's one issue I can't seem to
overcome. It might be my lack of BGP knowledge, but it seems like I'm
doing everything right. Here's the scenario:
$peer1 = cisco -- we're just sending full routes to $peer2 (not sure if my
coworker set this up as a reflector or what, but it works)
$peer2 = openbgpd on freebsd. I am successfully receiving full routes (8
seconds from a cold start. Amazing!)
$peer3 = openbgpd on openbsd. I'm trying to get all those routes from
$peer2 so we can test things between the openbgpd instances, but this is
not working.
The attached censored configs are the currents state of affairs and I'm
quite frustrated. I've tried so many different things. Google isn't being
kind to me.
Thank you for your time.
Regards,
Mark
#macros
peer1="X.X.0.1"
peer3="X.X.7.201"
#local_ip="X.X.7.202"
as="0000"
# global configuration
AS $as
log updates
# neighbors and peers
neighbor $peer1 {
remote-as $as
#local-address $local_ip
descr "router01.excelsior"
announce none
}
neighbor $peer3 {
remote-as $as
#local-address $local_ip
descr "openbsd-201"
announce all
route-reflector
}
# filters
allow from any
allow to any
#macros
peer1="X.X.7.202"
local_ip="X.X.7.201"
as="0000"
# global configuration
AS $as
log updates
# neighbors and peers
neighbor $peer1 {
remote-as $as
descr "freebsd-202"
announce none
}
# filters
allow from any
allow to any
