Re: [Openv6] APONF (Application-based Policy for Network Functions): new BOF for IETF 90 (requesting comments)

Hi Brian,

In addition to what Cathy answered I would like to mention that yes, it will be good to meet in Toronto and discuss the issues that you mentioned!

Best regards,
Georgios

________________________________________
Van: Zhouqian (Cathy) [cathy.zhou@huawei.com]
Verzonden: dinsdag 10 juni 2014 5:03
Aan: Brian E Carpenter; Karagiannis, G. (EWI)
CC: openv6@ietf.org
Onderwerp: RE: [Openv6] APONF (Application-based Policy for Network Functions): new BOF for IETF 90 (requesting comments)

Hi Brian,

Best Regards,
Cathy

> -----Original Message-----
> From: Openv6 [mailto:openv6-bounces@ietf.org] On Behalf Of Brian E
> Carpenter
> Sent: Tuesday, June 10, 2014 4:37 AM
> To: karagian@cs.utwente.nl
> Cc: openv6@ietf.org
> Subject: Re: [Openv6] APONF (Application-based Policy for Network Functions):
> new BOF for IETF 90 (requesting comments)
>
> Hi Georgios,
>
> On 08/06/2014 19:46, karagian@cs.utwente.nl wrote:
> > Hi Brian,
> >
> > Thanks for your comments!
> >
> > Regarding the relation to openv6, please note that APONF can be used to
> support an operator on deploying and configuring (and dynamically
> reconfiguring) different IPv4/IPv6 transition solutions depending on the applied
> scenarios and the expected user experience.
>
> Yes, that would be one application but I am still confused. I have seen no other
> traffic on the openv6 list, so what are we doing here? Why isn't this discussion
> on an APONF list, and where is the actual discussion of openv6?

Openv6 is one use case of APONF. We already had two side meetings of openv6 in both IETF 88 and IETF 89.
So we hope to use the expertise in this list to discuss APONF first.

>
> > Regarding the relation of APONF to UCAN, to my understanding, the UCAN
> BOF intends to expose several use cases that are focusing on Autonomic
> Networking (AN), i.e., self management.
> > APONF is focusing not focusing on solving AN issues. Please note that the
> dynamic (re)configuration issues mentioned in the APONF BOF (pre-charter) are
> not associated with self-management.
>
> Actually, any scenario that currently relies on a policy/management function
> and some kind of automation is a possible use case for autonomic networking. I
> don't see any of the actions involved in APONF as being out of scope for AN.
> Application-based policy intent would be considered as an input to the AN
> mechanisms, and the topics that you have listed as "network management and
> traffic policies"
> are exactly the type of objectives we would consider as targets for autonomic
> negotiation between network elements.
>
> The "the application-based policy protocol(s), mechanisms and models"
> that you mention might be exactly what is needed to communicate between
> applications and the AN policy mechanism.
APONF focuses on auto-mapping application's demand on network to the policy that can be understood by network. The applications could be the following ones, e.g., Traffic Engineering, Anti-DDoS, IPv6 transition, etc. These applications are not in the scope of Autonomic Network. One of APONF's objective is to define a set of models for such applications. These models may be used for AN or any other policy based use cases.

> If both BOFs are approved, we need to discuss in Toronto!
>
> Regards,
>    Brian
>
> >
> > Best regards,
> > Georgios
> >
> >
> > ________________________________________
> > Van: Brian E Carpenter [brian.e.carpenter@gmail.com]
> > Verzonden: zaterdag 7 juni 2014 22:39
> > Aan: Karagiannis, G. (EWI)
> > CC: openv6@ietf.org
> > Onderwerp: Re: [Openv6] APONF (Application-based Policy for Network
> > Functions): new BOF for IETF 90 (requesting comments)
> >
> > Hi,
> >
> > Can you explain the specific relationship between this BOF and IPv6?
> > As far as I can see, it is 100% independent of the IP version (and
> > that is a good thing).
> >
> > By the way, the topic of autonomic configuration by negotiation
> > between various network elements, which could include the elements
> > involved in the APONF scenarios, is the subject of another BOF
> > proposal, which also has no specific dependency on IPv6:
> > http://trac.tools.ietf.org/bof/trac/wiki/WikiStart#UCANproposal
> > APONF might be an additional use case for UCAN.
> >
> > Regards
> >    Brian Carpenter
> >
> > On 08/06/2014 06:01, karagian@cs.utwente.nl wrote:
> >> Hi all,
> >>
> >> Please note that a request for a BOF associated with Openv6 has been
> requested for IETF 90.
> >>
> >> The name of this BOF is:
> >>
> >> APONF (Application-based Policy for Network Functions).
> >>
> >> The main goal of APONF is to specify the application-based policy protocol(s),
> mechanisms and models required by transport applications to easily, accurately,
> and efficiently select and use the available communication network capabilities,
> i.e., network management and/or traffic policies.
> >>
> >> For more details on APONF, please see:
> >> http://trac.tools.ietf.org/bof/trac/#
> >>
> >> Please note that in association with this BOF two Internet drafts have been
> (up to now) submitted .
> >> These drafts are:
> >>
> >> APONF problem statement:
> >> http://tools.ietf.org/id/draft-karagiannis-aponf-problem-statement-00
> >> .txt
> >>
> >> APONF architecture:
> >> http://tools.ietf.org/id/draft-zhou-aponf-architecture-00.txt
> >>
> >> Your comments on the APONF BOF, and on the Internet drafts listed above
> are very much appreciated!
> >>
> >> Best regards,
> >> Georgios Karagiannis
> >>
> >>
> >>
> >> ---------------------------------------------------------------------
> >> ---
> >>
> >> _______________________________________________
> >> Openv6 mailing list
> >> Openv6@ietf.org
> >> https://www.ietf.org/mailman/listinfo/openv6
>
> _______________________________________________
> Openv6 mailing list
> Openv6@ietf.org
> https://www.ietf.org/mailman/listinfo/openv6