Re: [Banana] FW: New Version Notification for draft-leymann-banana-ecn-00.txt

Hi Michael

Thanks for your comments. Please see in line.

Lihao

> -----Original Message-----
> From: Michael Menth [mailto:menth@uni-tuebingen.de]
> Sent: Wednesday, September 13, 2017 4:31 PM
> To: chenlihao; banana@ietf.org
> Cc: Andreas Stockmayer; Tobias Scherer
> Subject: Re: [Banana] FW: New Version Notification for
> draft-leymann-banana-ecn-00.txt
> 
> Dear Lihao and coauthors,
> 
> thanks for your contribution. I have some comments.
> 
> 1) The relation of your document to
> https://www.ietf.org/id/draft-ietf-tsvwg-tunnel-congestion-feedback-05.txt
> may be clarified. That draft sends aggregated ECN feedback from a receiver to
> a controller using IPFIX while your proposal sends per-packet feedback on the
> tunnel to the tunnel ingress (which may be a controller).
>
Thanks for the information.

> 2) It may be helpful to extend the feedback information by packet loss.
> A generalization of your approach seems helpful.
> 
> 3) As 1) shows, feedback may be sent either in regular intervals or per-packet.
> That pertains for both congestion signals: loss and delay.
> Those are design options for general congestion feedback in BANANA.
> 
For 2) and 3), I agree that those are design options for general congestion feedback in BANANA and would be well defined in the future. Therefore, ECN notification could be supported by the general congestion feedback method of BANANA.

> 4) When using ECN information from the tunnel, there may be a double
> reaction to congestion: on the one hand traffic is shifted to another tunnel, on
> the other hand end-to-end flows reduce their rate. I don't say this is a problem,
> but we should be aware of it.
> 
Yes, we should be aware of it. And I think for BANANA scenarios, if the problem of congestion could be handled by shifting traffic to another tunnel, we might not need to reduce end-to-end flow rate.

> Kind regards,
> 
> Michael
> 
> 
> Am 12.09.2017 um 10:03 schrieb chenlihao:
> >
> > Hi all,
> >
> > The BANANA-ECN issue was discussed during the BANANA side meeting in
> Prague. We think that the idea of ECN operations for bonding tunnels make
> sense. Therefore, we post the draft "BANdwidth Aggregation for interNet
> Access (BANANA) ECN Operations for Bonding Tunnels".
> >
> > Please find the links of the draft below. Comments are welcomed!
> >
> >
> > Best regards,
> > Lihao
> >
> >
> > -----Original Message-----
> > From: internet-drafts@ietf.org [mailto:internet-drafts@ietf.org]
> > Sent: Tuesday, September 12, 2017 2:49 PM
> > To: Gengxuesong (Geng Xuesong); chenlihao; Zhangmingui (Martin); 4c
> > 67; Cornelius Heidemann; Liang Geng; Jun Shen; Nicolai Leymann
> > Subject: New Version Notification for draft-leymann-banana-ecn-00.txt
> >
> >
> > A new version of I-D, draft-leymann-banana-ecn-00.txt has been successfully
> submitted by Lihao Chen and posted to the IETF repository.
> >
> > Name:		draft-leymann-banana-ecn
> > Revision:	00
> > Title:		BANdwidth Aggregation for interNet Access (BANANA) ECN
> Operations for Bonding Tunnels
> > Document date:	2017-09-12
> > Group:		Individual Submission
> > Pages:		9
> > URL:
> https://www.ietf.org/internet-drafts/draft-leymann-banana-ecn-00.txt
> > Status:
> https://datatracker.ietf.org/doc/draft-leymann-banana-ecn/
> > Htmlized:       https://tools.ietf.org/html/draft-leymann-banana-ecn-00
> > Htmlized:
> https://datatracker.ietf.org/doc/html/draft-leymann-banana-ecn-00
> >
> >
> > Abstract:
> >    This document specifies a Bonding Tunnel ECN Mechanism that uses
> >    Explicit Congestion Notification (ECN) in bonding tunnels to notify
> >    congestion of a tunnel so that the load-balancing strategy of the
> >    tunnel ingress can be adjusted accordingly. Attributes for the
> >    control protocol of BANANA are defined to support this mechanism.
> >
> >
> >
> >
> > Please note that it may take a couple of minutes from the time of submission
> until the htmlized version and diff are available at tools.ietf.org.
> >
> > The IETF Secretariat
> >
> > _______________________________________________
> > Banana mailing list
> > Banana@ietf.org
> > https://www.ietf.org/mailman/listinfo/banana
> >
> 
> --
> Prof. Dr. habil. Michael Menth
> University of Tuebingen
> Faculty of Science
> Department of Computer Science
> Chair of Communication Networks
> Sand 13, 72076 Tuebingen, Germany
> phone: (+49)-7071/29-70505
> fax: (+49)-7071/29-5220
> mailto:menth@uni-tuebingen.de
> http://kn.inf.uni-tuebingen.de