Minutes IETF118: dmm: Mon 12:00
minutes-118-dmm-202311061200-01
Meeting Minutes | Distributed Mobility Management (dmm) WG | |
---|---|---|
Date and time | 2023-11-06 12:00 | |
Title | Minutes IETF118: dmm: Mon 12:00 | |
State | Active | |
Other versions | markdown | |
Last updated | 2023-11-12 |
minutes-118-dmm-202311061200-01
WG status update
Adaption call
(1) Draft-kohno-dmm-srv6mob-arch draft
- no significant issue
(2) mup-evaluation draft
- (Tianji) Request to post the issue in detail to DMM mailing.
Presentation
1. Mobility Aware Transport Network Slicing for 5G
- draft-ietf-dmm-tn-aware-mobility, John Kaippallimalil, 15 min.
Question/Comment:
-
(Chair)
-
IETF has a lot of transport technologies. Why still this draft
is stating ppr as the transport? -
Do you have any implementation of the udp source mapping for
slicing? - No. -
Need to work with the transport experts.
-
-
(Takahiro)
- How about c-plane slice? - the draft text covers the slicing of
the control plane as well.
- How about c-plane slice? - the draft text covers the slicing of
-
(Sudesh)
- Is the private network covered? - The draft could be covering
both private and public network. But having the discussion
offline to figure out if there are any issues.
- Is the private network covered? - The draft could be covering
Computing aware traffic steering use case of mobile user plane
- draft-duongph-dmm-computing-aware-ts-mup-sr, Dương Phùng Hà,
10min.
Question/Comment:
-
(John)
- Now looks like this is trying to combine based on the reverse
manner.
- Now looks like this is trying to combine based on the reverse
-
(Tom)
- What is the requirement on the data path? Also, what is the
impact on 5GC? What is the expectation to 5GC? — No intention to
make any change in 5GC
- What is the requirement on the data path? Also, what is the
-
(Tianji)
- CATS-MUP-C could be hacked to 5GC architecture to get the UE
location, etc which are handled by AMF, etc.. This is requesting
a lot of things to 5GC...
- CATS-MUP-C could be hacked to 5GC architecture to get the UE
-
(Yuya)
- Need to clarify the anycast service pdu session in detail...
-
(Hannu)
- Why this draft is submitted in DMM? Because this is not related
to “mobility”
- Why this draft is submitted in DMM? Because this is not related
-
(Marco)
- This draft can introduce the confusion to 5GC/CATS discussion.
The path/traffic steering should be independent from particular
mobility system.
- This draft can introduce the confusion to 5GC/CATS discussion.
-
(Arashmid)
- The draft assumes to use SRv6 but any operators should pick up
any transport technologies like SRv6, MPLS, etc…
- The draft assumes to use SRv6 but any operators should pick up
-
(Chair)
- How to advertise the route awarding CATS? - Route is advertised
with CATS metric (Need to describe it in the draft)
- How to advertise the route awarding CATS? - Route is advertised
-
(Tianji)
- Why SR? it seems that SR is outhogonal to what you propose in
the draft.
- Why SR? it seems that SR is outhogonal to what you propose in
-
(John)
- You don't need interwork with 3GPP if DNS resolution for that
service gives the anycast IP address.
- You don't need interwork with 3GPP if DNS resolution for that
-
(Xia Chen)
- Does 5G interface have a parameter to carry the Service-ID? 5GC
can create a pdu session per slice but this draft requires a pdu
session per service id. Is this function supported in 5GC? ==>
The author will discuss later…
- Does 5G interface have a parameter to carry the Service-ID? 5GC
SRH Reduction for SRv6 End.M.GTP6.E
- draft-kawakami-dmm-srv6-gtp6e-reduced, Yuya Kawakami, 20min.
Question/Comment:
-
(Tom)
- Is this efficient rather than looking up SRH?
If removing SRH, this is same as regular IPv6. So, Need to
clarify.
- Is this efficient rather than looking up SRH?
-
(Suresh)
- How does adressing works for longer prefix, what does protocol
use to send address? - Interwork segment route and session
transform route-1 of MUP-SAFI route are used to advertise the
address and SID.
- How does adressing works for longer prefix, what does protocol