Folks,
If any of you BGP peer with DREN, as we do, you will be getting
(as part of their AS668) a couple of subnets of 147.73/16:
B 147.240.0.0/16 [5/0] via 138.18.169.1, 01:44:58
147.73.0.0/16 is variably subnetted, 2 subnets, 2
B 147.73.234.0/28 [5/0] via 138.18.169.1, 01:40:24
B 147.73.234.128/27 [5/0] via 138.18.169.1, 01:39:54
This could cause customers visiting SC97 to complain about
connectivity to their local network from SC97 unless you are also
receiving routing for the remaining 147.73 space from some other
peer.
I thought some subset of the nanog population might want to know
since it caused me a bit of trouble shooting this morning, and nobody
at DREN is available.
Phil
Phil,
The subnets you are receiving via your BGP session with DREN support the DoD
HPCMP portions of the net at SC97. There is a DREN connection to SC97 to
support the demonstrations. Therefore you should reach the DoD HPCMP
resources at SC97 via your DREN connection, and the rest of your SC97 traffic
should follow the less explicit 147.73/16 path. DREN is not announcing
these subnets to other peers at the exchanges, only to DREN members.
Meanwhile, in response to your message, Tom Kile installed a filter at PLK
so you should no longer receive these prefixes via DREN.
--Ron
At SC97, we saw 147.73/16 injected into the vBNS yesterday by DREN.
this caused vBNS sites to traverse the vBNS to SDSC then hop to DREN
to get to SJCC while the vBNS/NTON link into SJ was being worked on.
this was quite unexpected.
lw
Phil,
The subnets you are receiving via your BGP session with DREN support the
DoD
HPCMP portions of the net at SC97. There is a DREN connection to SC97 to
support the demonstrations. Therefore you should reach the DoD HPCMP
resources at SC97 via your DREN connection, and the rest of your SC97
traffic