Minutes IETF101: nfvrg

Meeting Minutes Network Function Virtualization (nfvrg) RG
Title Minutes IETF101: nfvrg
State Active
Other versions plain text
Last updated 2018-04-09

Meeting Minutes

NFVRG Meeting Wed March 21 2018 1330-1500
Room Sandringham

Note takers: Linda Dunbar, Pedro Aranda
Jabber scribe: Evangelos Haleplidis

Update on current and proposed drafts ~ 15 min
got good review
Document is ready
Diego is going to ask to put the document in the RFC queue

No comments from the audience

solution enabling network based discovery and advisement of VIM and NFVO
[Linda Dunbar] who do you advise the VIM and NFVO?
Answer: advise to users, such as Compute/Storage/Network resources to clients
Csrlos: looking for resources, so someone may advertise the resources (s)he
might be willing to offer

Diego: we can ask operators with operation experience to review and comment.
Incorportate experience from the MAMI experiments

Introducing NFV-based experimental facilities ~ 30 min
5GINFIRE (A. Eisenmann, University of Bristol)
Yokov Stein: the use case come to my mind is changing split.
The ideal of 5G phy
Yakov: where do you instantiate the NVF?
Answers: we provide the physical servers to run the NFVs
Tin: from National Research Lab, would like to help.
Kostas ?? how to achieve the slice?
Kostas?? How do you achieve the virtulation (isolation)?
Answer: Manually
University: do you have flexibility on what VPN in use? can you only select the
pre-stored NFV? L3VPN Christos: Looks like MEC Do you want to extend it to a
MEC model Answer: yes Christos: Enhancements for the operational capability

SOFTFIRE (S. Vural, 5GIC - University of Surrey)
brief the Near finished EU project
chair: are you going to publish the result?
Answer: in couple of months. OpenBATON is already available
Al Morton (ATT): what kind of information will you benchmark?
 Answer: this is a very important information. Software match with physical.
Peter: the results are not published yet?
Answer: 2 white paper would publish soon.
Q: is it over after next Hackthone?
A: will transfer knowloedge to next projects
Q: IP v6 running / problems you ran into
A: out of scope

NFV in Rich Communication Services (Y. Stein, RAD) ~ 30 min
QoE is subjective measurement, can't be measured
QoS is measurable, but sometimes doesn't mean anything to end users

University: stronngly disgree you slide on Link Down improving the QoE
Ross: If you take into the account the packets being added or removed, it would
get more accurate QoE Al Morton (ATT): seperate measurement. Yokov: we don't
know explict equation.