Re: [AVT] Congestion is unidirectional or bidirectional ???

["James M. Polk" <jmpolk at cisco.com>]

Congestion *should*, from a network point of view, be considered an 
outbound interface problem, therefore the example below should support the 
claim that it is unidirectional. Even if both directions of a bidirectional 
flow traverse the same intermediate node (i.e. router), the outbound 
interface is likely different for each flow.

In the cases of hairpinning on a router, or in and out of the network 
attachment of a node to the network, one flow can cause congestion of 
another, or all other flows, on that same interface. This could be the case 
when looking at a server.  However, the inbound traffic should not cause 
congestion on the outbound traffic directly, if that server is full duplex 
attached to the network.

At 04:21 PM 11/1/2005 +0100, Saverio Mascolo wrote:
> i would say it is unnidirectional for the flow that goes from A to B, but 
> this flow could be affected by congestion on the reverse path in case the 
> flow A-->B is ack clocked such as in the case of TCP
>
> savio
> ----- Original Message -----
> From: <mailto:aco1967 at gmail.com>Alejandro Cabrera Obed
> To: <mailto:avt at ietf.org>AVT Lista
> Sent: Tuesday, November 01, 2005 2:23 PM
> Subject: [AVT] Congestion is unidirectional or bidirectional ???
>
> Dear all,
>
> I have to ask you because I couldn't find the answer on the web to my 
> question...sorry again.
>
> Suppose I transmit a media flow between A and B, and these points are 
> connected through Internet2 or a similar high-speed network.
>
> Suppose that in a given moment, there is a congestion in the path from A 
> to B (let's say caused by a given router) that makes that no packets reach B.
>
> At this same moment........is there a congestion from B to A too ???
>
> Congestion states are a unidirectional or a bidirectional problem ???
>
> One person from Internet2 tells me that he supposes and has experienced 
> that congestion is mostly unidirectional, but he has not a good academic 
> cite to confirm this.
>
> Thanks a lot,
>
> Alejandro
>
>
>
> --
> Alejandro Cabrera Obed
> <mailto:aco1967 at gmail.com>aco1967 at gmail.com
>
>
> ----------
> _______________________________________________
> Audio/Video Transport Working Group
> avt at ietf.org
> https://www1.ietf.org/mailman/listinfo/avt
>
> _______________________________________________
> Audio/Video Transport Working Group
> avt at ietf.org
> https://www1.ietf.org/mailman/listinfo/avt


cheers,
James

                                *******************
                Truth is not to be argued... it is to be presented.

_______________________________________________
Audio/Video Transport Working Group
avt at ietf.org
https://www1.ietf.org/mailman/listinfo/avt