From: Yuval Lifshitz <ylifshitz@sandvine.com>
To: "Bertz, Lyle T [CTO]" <Lyle.T.Bertz@sprint.com>, Dave Dolson
 <ddolson@sandvine.com>, "dime@ietf.org" <dime@ietf.org>
Thread-Topic: New I-D: draft-bertz-dime-perflowmbr-00
Thread-Index: AdHXoI27HhmJQL9SSVCCcgs9eQDIkwABGo+QAAAmleAAABKD8AAAFziQAAE97GA=
Date: Wed, 6 Jul 2016 17:30:17 +0000
Message-ID: <C43C255C7106314F8D13D03FA20CFE4930C8AE19@wtl-exchp-2.sandvine.com>
References: <bda537d9971b46858976bb40b11ef23c@PLSWE13M07.ad.sprint.com>
 <E8355113905631478EFF04F5AA706E9830FE6E7F@wtl-exchp-2.sandvine.com>
 <d5a2213039d44096b87253727dbb8d6a@PLSWE13M07.ad.sprint.com>
 <E8355113905631478EFF04F5AA706E9830FE6F64@wtl-exchp-2.sandvine.com>
 <5976352e235d43598a2f2d86c919a65c@PLSWE13M07.ad.sprint.com>
In-Reply-To: <5976352e235d43598a2f2d86c919a65c@PLSWE13M07.ad.sprint.com>
Accept-Language: en-CA, en-US
Content-Language: en-US
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Archived-At: <https://mailarchive.ietf.org/arch/msg/dime/C1ZbajDAxGZX0tES4cUHVQYgQnM>
Subject: Re: [Dime] New I-D: draft-bertz-dime-perflowmbr-00
Precedence: list

* maybe good to mention how these rates work with the overall rates defined=
 for all flows of the filter
* types in section 4 should be Unsigned32=20

From: DiME [mailto:dime-bounces@ietf.org] On Behalf Of Bertz, Lyle T [CTO]
Sent: Wednesday, July 06, 2016 7:55 PM
To: Dave Dolson; dime@ietf.org
Subject: Re: [Dime] New I-D: draft-bertz-dime-perflowmbr-00

Yes, if you are using the IPFilterRule or RFC 5777 Filter-Rule as it is tie=
d to the Protocol field.=A0 I also intended for this not to conflict with m=
ulti-path based protocols.=A0=20

If you are applying these attributes to another kind of Rule, e.g. 3GPP ADC=
 Rule, then sticking to the L3 is still fine.=A0 However, if you want to us=
e it for the Application level flow this document would need to be changed =
and we'd need to incorporate it in the draft.=A0 Is that something I should=
 look at for this document.=A0 More work would be needed to clarify interac=
tion with multipath and other protocols though.


From: Dave Dolson [mailto:ddolson@sandvine.com]=20
Sent: Wednesday, July 06, 2016 11:48 AM
To: Bertz, Lyle T [CTO] <Lyle.T.Bertz@sprint.com>; dime@ietf.org
Subject: RE: New I-D: draft-bertz-dime-perflowmbr-00

OK, I see in Security Considerations.

For your use-case, is it acceptable that a client could open two 5-tuple co=
nnections to achieve double the rate?




From: Bertz, Lyle T [CTO] [mailto:Lyle.T.Bertz@sprint.com]=20
Sent: Wednesday, July 06, 2016 12:44 PM
To: Dave Dolson; dime@ietf.org
Subject: RE: New I-D: draft-bertz-dime-perflowmbr-00

Hmm.

Should say IP flow (5-tuple) but that may be in the wrong location.


From: Dave Dolson [mailto:ddolson@sandvine.com]=20
Sent: Wednesday, July 06, 2016 11:43 AM
To: Bertz, Lyle T [CTO] <Lyle.T.Bertz@sprint.com>; dime@ietf.org
Subject: RE: New I-D: draft-bertz-dime-perflowmbr-00

Lyle,
This document doesn't define a "flow". What did you have in mind as a defin=
ition?

It sounds like you are thinking of a TCP or UDP connection? So if the inten=
t were to limit video rates, each TCP or UDP connection would be granted th=
e specified rate?

-Dave


From: DiME [mailto:dime-bounces@ietf.org] On Behalf Of Bertz, Lyle T [CTO]
Sent: Wednesday, July 06, 2016 12:31 PM
To: dime@ietf.org
Subject: [Dime] New I-D: draft-bertz-dime-perflowmbr-00

All,=20

I have added a new I-D for consideration.=A0 It permits setting *per flow* =
maximum bit rates.=A0 This can serve in some applications, e.g. video throt=
tling, to limit features such as resolutions as opposed to requiring deeper=
 inspection.

https://datatracker.ietf.org/doc/draft-bertz-dime-perflowmbr/

Thank you.

Lyle

________________________________________
Learn more on how to switch to Sprint and save 50% on most Verizon, AT&T or=
 T-Mobile rates. See sprint.com/50off for details.=20
________________________________________

This e-mail may contain Sprint proprietary information intended for the sol=
e use of the recipient(s). Any use by others is prohibited. If you are not =
the intended recipient, please contact the sender and delete all copies of =
the message.