[36376] in North American Network Operators' Group
RE: Faster 'Net growth rate raises fears about routers
daemon@ATHENA.MIT.EDU (Tony Barber)
Wed Apr 4 05:23:02 2001
Message-Id: <3.0.6.32.20010404101900.0098d430@pop.ukgateway.net>
Date: Wed, 04 Apr 2001 10:19:00 +0100
To: nanog@merit.edu
From: Tony Barber <acb@ukgateway.net>
In-Reply-To: <9DC8BBAD4FF100408FC7D18D1F092286039DC2@condor.mhsc.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Errors-To: owner-nanog-outgoing@merit.edu
At 10:55 AM 03/04/01 -0700, Roeland Meyer wrote:
>
>> From: Travis Pugh [mailto:tpugh@shore.net]
>> Sent: Tuesday, April 03, 2001 10:34 AM
>
>> I'm at a multi-POP network in Boston. We've had great luck selling
>> customers a Verizon circuit into one of our POPs and a
>> Worldcom circuit
>> into a different one. It costs more, but they don't have nearly the
>> exposure of a single circuit customer. However, if you're
>> not set up to
>> do this, the appropriate level of paranoia calls for circuits to two
>> different providers. Maybe if SPs really addressed availability
>> requirements of their customers, it wouldn't be such an issue.
>
>The problem with this, if done, is that we back right into the other problem
>of prefix filtering. If the customer has a /19 or /20, there is generally no
>problem. But, if it is the usual case (/24) then only one of the upstreams
>can aggragate the routes up. What is the other ISP to do? How would this be
>made to work? BTW, this is exactly the reason we weren't fully multi-homed
>yet.
>
Cisco has a knob for conditional advertising. If this functionality were
standardised, documented and marketed more installation consultants could
make use of it. This would undoubtably help the aggregation cause as most of
the time fault conditions would not be active.
Tony
