Minutes interim-2023-alto-02: Tue 14:00
minutes-interim-2023-alto-02-202304111400-00
Meeting Minutes | Application-Layer Traffic Optimization (alto) WG | |
---|---|---|
Date and time | 2023-04-11 14:00 | |
Title | Minutes interim-2023-alto-02: Tue 14:00 | |
State | Active | |
Other versions | markdown | |
Last updated | 2023-04-12 |
minutes-interim-2023-alto-02-202304111400-00
ALTO Interim #2, April 11, 2024
Chairs: Mohamed Boucadair & Qin Wu
Minutes: TBC
Introduction (5 mn)
Chairs go through Agenda
OAM WGLC Status (25 mn)
- JZ goes through list of issues and PRs
- We have addressed yangdoctor reviews
-
JZ goes into the resolved issues and explains resolution
- E.g.: new typedefs added
-
RY asks about rationale for pattern regex
- There is no standard for the pattern
- Look at RFC 7285 PID definition
-
Changed the fixed reporting interval to a configured parameter
-
Encoding of meta-value
- QW share this issue to the mailing list
-
JZ will write to Andy right after interim
-
When will you issue new version?
- Tomorrow
-
QW address all yangdoctors reviews and issue another review
iteration
New Transport WGLC Status (25 mn)
- Version 7, working on working 8, in github, not submitted to
datatracker yet - Push order does not need to be strictly in order
- Security comments resolved and reviewer confirmed his comments are
resolved - OPSDIR mostly nits. Adding new references in next revision.
-
TSVART:
- Transport requirements not laid out
- Backward compatibility not stated
- Increment between updates be 1?
- More efficient encoding
-
HTTPDIR blocker:
-
HTTP 2/3 server push
- lack of implementation support
- not the intended use of server push
- Two options:
- Split HTTP2/3 docs
- Make the case that server push is justified
- Server push not widely used, so if it goes away that's
an issue - Main benefit is reduces round trip delay
- MB what are we losing if we don't have server push
- RY we are going to explain the motivation of push
- Follow up with Martin and Spencer
- Verify the future of server push in HTTP
-
Session bounded to single connection, should be based on
requests- QW Bring discussion to mailing list
-
Meta layer too complex
-
Misc.
WG Future Discussion (Reminder)
Next Interims
- Interim #3: Deployment Catalysts
- Reuse existing network resources to identify ALTO resources
(e.g., BGP communities) - Proposals to fix the integration complexity and integration with
data sources
- Reuse existing network resources to identify ALTO resources
- Interim #4: OAM/Transport I-Ds
- Interim #5: Deployment Catalysts
- Security/privacy isolation