RE: [PCN] traffic matrix scenario

"Geib, Ruediger" <> Wed, 31 October 2007 11:13 UTC

Return-path: <>
Received: from [] ( by with esmtp (Exim 4.43) id 1InBVc-00034S-QW; Wed, 31 Oct 2007 07:13:16 -0400
Received: from pcn by with local (Exim 4.43) id 1InBVb-00033v-3a for; Wed, 31 Oct 2007 07:13:15 -0400
Received: from [] ( by with esmtp (Exim 4.43) id 1InBVa-000331-GK for; Wed, 31 Oct 2007 07:13:14 -0400
Received: from ([]) by with esmtp (Exim 4.43) id 1InBVU-0007Pv-Ur for; Wed, 31 Oct 2007 07:13:14 -0400
Received: from by with ESMTP; Wed, 31 Oct 2007 12:12:59 +0100
Received: from ([]) by with Microsoft SMTPSVC(6.0.3790.3959); Wed, 31 Oct 2007 12:12:59 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [PCN] traffic matrix scenario
Date: Wed, 31 Oct 2007 12:12:58 +0100
Message-Id: <>
In-Reply-To: <>
Thread-Topic: traffic matrix scenario
Thread-Index: AcgbA+IAhc9rfEAcTjSuKU0GO02DLwAlhhvw
From: "Geib, Ruediger" <>
X-OriginalArrivalTime: 31 Oct 2007 11:12:59.0562 (UTC) FILETIME=[FEF270A0:01C81BAE]
X-Spam-Score: -1.0 (-)
X-Scan-Signature: 6e922792024732fb1bb6f346e63517e4
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: PCN WG list <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>


I'd like to return to my point, that probing is only required if there's
any pre-congestion in the network.

The question now is, what's the probability, that two end points try to
set up a first flow between them and none of both is aware of
pre-congestion, which is already on one of the links to be crossed by
that flow. The congestion isn't on the boundary to interior links (then
either ingress or egress node receive pre-congestion marked packets or
receive feedback on them). Hence you look for a pre-congested interior
link which isn't crossed by any flow to or from the boundary nodes
trying to set up a new flow. As this link is pre-congested, there's a
heavy demand between many other end points, but the link is not used by
a single flow of the end points trying to set up the new flow. That's
the situation when you may admit a flow on a link which is already in
pre congestion mode. 

As mentioned in my other mail, I don't think we should try to measure
whether a new flow would push an aggregate to pre-congestion mode. I
regard this case as standard operation, the decision should always be:



-----Original Message-----
From: []
Sent: Tuesday, October 30, 2007 3:48 PM
Subject: [PCN] traffic matrix scenario


Ben & I have made some estimates, based on a future busy hour for a
national network with about 100 PCN-boundary-nodes. We assume that the
case of interest is where the traffic follows this normal pattern and
that the network is (becoming) pre-congested due to a massive surge of
traffic outside this traffic pattern. 

Our estimate is that 3/4 of ingress-egress-aggregates will have traffic
of less than 1.5 Erlang. The distribution is long-tailed, for example
there are several ingress-egress-aggregates with > 4000 Erlangs.

We have made some speculations about how this translates into traffic on
particular interfaces. The topology of the network has each
PCN-boundary-node attached to 2 different PCN-interior-nodes;
PCN-interior-nodes are more interconnected (> dual-homed). 
So for a particular PCN-boundary-node to PCN-interior-node interface, it
potentially has 50 ingress-egress-aggregates on it; we estimate that
actually 25 of these will not have a flow, 10 will have 1 and 15 will
have >1.
For a PCN-interior-node to PCN-interior-node interface, the topology
suggests potentially 300 ingress-egress-aggregates on it; the fractions
are the same as before, ie we estimate that actually 150 of these will
not have a flow, 60 will have 1 and 90 will have >1.

We re-iterate that these are very rough estimates. But they strongly
suggest that there are likely to be significant numbers of aggregates
with very few flows under nearly all circumstances.

Best wishes,
Phil & Ben
> -----Original Message-----
> From: Lars Eggert []
> Sent: 25 October 2007 13:25
> To: Eardley,PL,Philip,CXR9 R
> Cc:;;
> Subject: Re: [PCN] Architecture draft - probing section & general
> On 2007-10-25, at 13:56, ext wrote:
> > Lars, I believe the guess is that there'll be many ingress-egress 
> > pairs with none or very few flows at a particular moment. Not a 
> > random distribution of calls onto all possible pairs. Basically 
> > there are many calls between London & Manchester and very few 
> > between truro and Shetland.
> I can agree with that. But note that that alone is not a problem yet.
> The problem occurs only when several of these "empty" ingress pairs 
> share a bottleneck *and* when single flows start simultaneously 
> sending across them at a combined rate that will push the bottleneck 
> directly into overload.
> I don't believe this can be very common. Wouldn't you provision your 
> network such that interior routers could handle at least one flow per 
> ingress/egress pair? In which case there is no issue unless there's a 
> failure. And why not let flow termination handle these rare cases?
> > I & Ben will try and extract some numbers from the [bt] predicted 
> > future traffic matrix.
> That'd be *very* useful! We're all guessing probabilities here, and 
> data would help.
> Lars

PCN mailing list