Minutes interim-2023-mpls-06: Thu 14:00
minutes-interim-2023-mpls-06-202306291400-00
| Meeting Minutes | Multiprotocol Label Switching (mpls) WG | |
|---|---|---|
| Date and time | 2023-06-29 14:00 | |
| Title | Minutes interim-2023-mpls-06: Thu 14:00 | |
| State | Active | |
| Other versions | markdown | |
| Last updated | 2023-07-05 |
MPLS Interim meeting on 2023-06-29
The intention with this meeting is to see if it is possible to
converge furthr on the discussion on "Multiple ISD instances in
a packet" that we have had on the mailing list.
I have asked 3 of the more active participants in that discussion
to prepare one slide each focusing on two questions "What are the
issues" and "What could consensus look like".
If time allows we will also have a discussions around a slide set
from Jie - "MNA ISD Limitation Analysis".
Plese find the agenda below.
-
Agenda bashing
All -
Action Items
Tarek -
Multiple ISD instances in a packet
chairs
Joel presenting.
Jags: Do you need to copy the content or just indicate the existence of
the HBH MNA ISD?
Joel: You need to copy the content. Or you indicate the next HBH also
needs to be processed.
Matthew: Need to be careful about the Hop-by-Hop concept. which are the
hops for Hierarchical tunnels? Are the HBH MNAs applicable to the newly
added tunnel. There are cases where the outer H-b-H actions may not
apply to the tunneled nodes. There may be cases where H-b-H actions are
applicable e2e.
Joel: Hope this is not case by case.
Joel: avoid multiple choices which complicate solutions.
Jie: May need to cover different behavoirs in the base solution
document, with are required by different applications.
Joel: Don't want to specify multiple options, want to have just one
default behavior, and other behavior can be specified by different
applications.
Tony:
Joel:
Tony: An example: ingress insert slice ID, some transit node add
entropy, should both be last till the egress?
Joel: They can have different scope.
- MNA ISD Limitation Analysis
Jie
Jie: For legacy hardwares with limited capability, if we want them to
support some MNA functionality, the limitations of legacy hardware need
to be considered. While for applications which cannot be parsed or
processed by legacy hardware regardless where the data is carried in the
packet (ISD or PSD), closer to the top of stack would not be an
important aspect to consider, and other limitations of ISD (encoding,
processing) may need to be considered.
- AOB