Possible topic?

Hi there,
  Is it possible, even at this late date, to allow someone from MFS to
comment at NANOG on what is being done to relieve inter switch congestion
at MAE East and West and what kind of time frames we can expect a
resolution in?

This is seriously affecting the performance of the Internet as a whole, and
seems to me to be a very valid operational topic for the conference.

Justin W. Newton wrote:

Hi there,
        Is it possible, even at this late date, to allow someone from MFS to
comment at NANOG on what is being done to relieve inter switch congestion
at MAE East and West and what kind of time frames we can expect a
resolution in?

    I second that motion !

This is seriously affecting the performance of the Internet as a whole, and
seems to me to be a very valid operational topic for the conference.

  With about 1/3 of all global internet passing through that parking
garage out there, 20% loss starts to take on a whole new meaning...

  While we are queing it up..........

What is the status of the HOL at the GIGA's ? I understand the new
code went in....

But, I guess you still can't put 5 pounds of "stuff" in a 3
pound bag , huh ? :wink:

Seriously, the Nanog would be a great time for MAE-E/W to speak up.

Just curious, but how was that figure arrived at?

-dorian

==> With about 1/3 of all global internet passing through that parking
==>garage out there, 20% loss starts to take on a whole new meaning...

Looks like someone else saw the Data Communications hype.

/cah

Dorian R. Kim wrote:

> With about 1/3 of all global internet passing through that parking
> garage out there, 20% loss starts to take on a whole new meaning...

Just curious, but how was that figure arrived at?

   Fair enough, I am quoting the newspapers. They recently ran an
article about MAE-E, and made the claim
that nearly 1/3 of all internet passes through MAE-E. I did not question
this revelation...
Perhaps I should have.

   After all, it made it into print, it *can't* be wrong! :wink:

   Unless you mean the 20%, in which case it is pure conjecture. When
*I* see MAE-E go lossy, it is typically around 20%. I have no idea why
"The Loss" favors this number. Or, even if other carriers hover about
20% when MAE-E goes lossy.... Food for thought though........

    Cheers.

Just wondering through MAE graphs. It looks like around Oct 3rd Mae-East
was moving about 1300Mbit or so. After that its been unable to rise about
660Mbit or so.

From what WorldComm just told us, Scott Gross and DEC are working on a

software patch "around the clock" to solve the problem.

-Deepak.

==>On Tue, 21 Oct 1997, Richard Irving wrote:
==>
==>> With about 1/3 of all global internet passing through that parking
==>> garage out there, 20% loss starts to take on a whole new meaning...
==>
==>Just curious, but how was that figure arrived at?

This figure came out of the most recent Data Communications magazine. The
cover had a picture of the parking garage with a cadillac and some
grandmother driving it near a secured door. It had a dark, scary caption
"One-third of the internet's traffic runs through this underground parking
garage in McLean, VA. Should you be scared?" or something similar.

/cah

   Fair enough, I am quoting the newspapers. They recently ran an article
about MAE-E, and made the claim that nearly 1/3 of all internet passes
through MAE-E.

Maybe 1/3 of all traffic *tries* to go through MAE-East. :slight_smile:

randy

Then perhaps it might of interest to start a meta discussion about
problems of scalability of Gigaswitch based exchange points as well as
possible future direction of such large exchange points?

-dorian

AFAIK, there are only two types of exchanges out there today:

    1. Switched Ethernet and/or FDDI
    2. ATM

  We all know switched ethernet doesn't scale by itself,
  which is why we've gone to switched FDDI.

  And, it's quite possible that we're reaching the end of
  scalability for switched FDDI -- unless, of course, some
  amazing new technolgy comes along and fixes everything in
  one swell foop.
  
  Are the ATM-based NAP's having any problems with scalability
  these days? How far can they go?

  Personally, I'd also be really curious as to any testing
  that people have done or are doing in terms of other forms
  of NAP fabric. Will gigabit ethernet save us all?

Gigabit Ethernet as a backbone technology -- too little, too late.
IMO, betting on physical media speeds keeping up with traffic
is silly.

Actually, there is at least one known way to build an exchange
point which scales nearly indefinitely. We're working on making
it real. Other folks may have other ideas.

--vadim

Actually, there is at least one known way to build an exchange
point which scales nearly indefinitely. We're working on making
it real. Other folks may have other ideas.

  Cool, Vadim; welcome to the ATM bandwagon!

  :-)

  -alan

Alan Hannan wrote:

> Actually, there is at least one known way to build an exchange
> point which scales nearly indefinitely. We're working on making
> it real. Other folks may have other ideas.

  Cool, Vadim; welcome to the ATM bandwagon!

  :-)

Oh, yeah, sure, dude. :slight_smile:

--vadim

Is there a tour for this place?

  of NAP fabric. Will gigabit ethernet save us all?

I think the question is, "Will *any* technology ever be able to save us
all?" I think (IMHO, and I am no expert, of course, as I have been told so
many times) that the need will always be a step or two ahead of what is
available. Sad; but true.

Aren't all the Gigaswitches from DEC ATM backed anyway? (FDDI is an
emulation on an OC3/12 card slot?) That's what the DEC literature seemed
to say when I perused it last.

-Deepak.

I guess you mean 'only two types of _public_ exchanges'. The answer
we all come up with might very well be that exchanging traffic over
public media is just not a good idea. The network might very well
have gotten to the point that it just doesn't make sense to have such
a vital piece of it completely under a third parties control.

Alec

Aren't all the Gigaswitches from DEC ATM backed anyway? (FDDI is an
emulation on an OC3/12 card slot?) That's what the DEC literature seemed
to say when I perused it last.

-Deepak.

Absolutely not. The GS/F and the GS/A are not the same. They even employ
slightly different crossbar architectures. The GS/F does support an ATM
line card however.

ss

Then perhaps it might of interest to start a meta discussion about
problems of scalability of Gigaswitch based exchange points as well as
possible future direction of such large exchange points?

AFAIK, there are only two types of exchanges out there today:

  1. Switched Ethernet and/or FDDI
  2. ATM

We all know switched ethernet doesn't scale by itself,
which is why we've gone to switched FDDI.

And, it's quite possible that we're reaching the end of
scalability for switched FDDI -- unless, of course, some
amazing new technolgy comes along and fixes everything in
one swell foop.

FDDI is a finished standard and no more bandwidth upgrades are planned.
Ethernet, however, is now in the gigabit range. It may be extended to
operate in the multi-gigabit range, if the IEEE decides the frame format
and the media access method will work at such speeds. Products based on
GbE are in use today. A NAP can drop in one or two GbE switches, connect
them to existing GIGAswitch/FDDIs, and scale (somewhat) from there. Come
to NANOG this next week to see how this can be done.

The nice thing about the GbE switches (I note the Prominent Cajun P550 and
the DEC GIGAswitch/Ethernet switches) is that they give you instant access
to over 45 Gbps of (almost) non-blocking switch fabric. GigaLabs has a
monster of a switch with over 128 Gbps of non-blocking fabric (GbE uplinks
only however). The only downsides right now with most current GbE switches
is that they only support 10/100BASE and GbE uplinks, offer no QoS and
limited flow-control mechanisms. The flow-control protocol for GbE has not
yet been finalized by the IEEE. All implementations of this are
vendor-proprietary. (Be sure your vendor upgrades you to the Standard at
no cost!).

Connecting a router to a switch using a standard a Fast Ethernet adapter
gives you 200 Mbps of full-duplex throughput. This is adequate for anyone
with an OC-3 behind their router. Hell, anyone with more bandwidth into
the NAP can probably purchase multiple Fast Ethernet links from the NAP
operator. These adapters are less costly than full-duplex FDDI adapters,
and the prices for switch ports and router adapters are declining.

I think that GbE technology permits the NAPs to support increasing
bandwidth requirements for the foreseeable future. Beyond that, maybe the
ISPs with heavy-duty requirements should consider private interconnects
outside the NAPs.

Steven

Then perhaps it might of interest to start a meta discussion about
problems of scalability of Gigaswitch based exchange points as well as
possible future direction of such large exchange points?

The two primary questions to answer this are "What technology is readily
available?", and "What is the anticipated need?".

It would be nice if MFS gave historical data at
http://www.mfst.com/MAE/west.stats.html in a format like MRTG does, rather
than just listing weekly traffic ad naseum.

Already a factor of 10 capacity is available from DEC in their new Gigabit
ethernet Gigaswitches. The new 12 slot chasis supposedly will run around
$9500, and a two port Gigibit ethernet adapter will supposedly retail for
around $4500 (there also is a multiport 100 base T card). This is much
cheaper than the FDDI Gigaswitch pricing per port.

Further, yet another factor of 10 is supposedly around the corner from the
company Packet Engines, which has a 10 Gbps ethernet(!) card in
development for their gigabit ethernet switch.

Both of these developments, coupled with the fact that long distance
pricing appears to not be dropping much (and therefore limiting the amount
of bandwith brought to an exchange), should make the exchanges much more
likely to be able to handle to load.

+------------------- H U R R I C A N E - E L E C T R I C -------------------+