Michael: I'm a little perplexed; it seems that the outbound metric for our peering with AS 2699 has been 127 since at least 24 May (when I switched to peering directly with AS 2699, instead of the two fake ASs). I'm not aware of any change with the handling of the backbone route, and the gated binaries on the machines in question have not changed (see below). Steve W.: Do you know of any change which would have affected this? (The gated binaries on ENSSs 144 and 145 have not been changed in some time, so...?) Is 2699 now beering with EBGP as well as EGP (are there new sessions)? Thanks, Steve R.

Date: Fri, 23 Sep 1994 15:01:19 EDT From: "Michael A. Patton" <MAP@BBN.COM> To: skw@merit.edu, sjr@merit.edu CC: nwg@merit.edu, dsi-ops@dsi.net, dsi-routing@dsi.net

Steves: (I can never quite remember which of you is which :-)

Sometime during the night (I assume when the most recent update was done), the NSFnet peer sessions with the DSI changed. Specifically, the metric for 140.222 is now 127. This value prevents us from readvertising that number to any of our customer sites. Several of them are multi-homed and use that advertisement in doing their default route selection. Since, for that reason, it only affects customers who have alternate paths, it is not critical, but we would appreciate if you could correct it at your earliest convenience.

__ /| /| /| \ Michael A. Patton, Network Analyst / | / | /_|__/ Defense Simulation Internet / |/ |/ |atton Bolt Beranek and Newman