Re: nisc-01742 (ticket update) Re: Mail problems at OpenSystems, Inc.

Sean Doran <> Wed, 08 March 1995 21:04 UTC

Received: from by IETF.CNRI.Reston.VA.US id aa10754; 8 Mar 95 16:04 EST
Received: from CNRI.Reston.VA.US by IETF.CNRI.Reston.VA.US id aa10746; 8 Mar 95 16:04 EST
Received: from by CNRI.Reston.VA.US id aa13974; 8 Mar 95 16:04 EST
Received: from ( []) by (8.6.10/merit-2.0) with ESMTP id PAA00580 for <>; Wed, 8 Mar 1995 15:56:47 -0500
Received: (from smd@localhost) by (8.6.9/8.6.9) id PAA27846; Wed, 8 Mar 1995 15:56:44 -0500
Date: Wed, 8 Mar 1995 15:56:44 -0500
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Sean Doran <>
Message-Id: <>
Subject: Re: nisc-01742 (ticket update) Re: Mail problems at OpenSystems, Inc.

| The worst flap echoeing offender is undoubtly ANS -- we hear
| full routing table from them.

Just to clarify this so that the ANS gang don't feel
unfairly persecuted -

We don't do much inbound as-path filtering on AS 690,
and won't until ANS's cooperative agreement expires and
the NSFNET backbone service ends.

The reason for this is largely historical; the NSFNET
backbone service has traditionally been ICM's last-resort
backup to other networks (this kinda makes sense given that
ICM manages international connectivity to the NSFNET
backbone service), and as SprintLink traditionally has been
transited to the NSFNET backbone service by ICM, well, the
whole spaghetti mess involved having to take full routing
from ANS at the FIXes.

SprintLink doesn't take everything from ANS at the
Pennsauken NAP, although we will take even less after the
end of April when the NSFNET backbone service ends...

Once that happens, I imagine that ANS will stop sending
us other people's routes altogether, possibly before we
get an equivalent inbound filter in place on our side.
This will eliminate AS 690 as the great big route-flap
echo chamber.

The transition is lots of fun. :-)