Skip to main content

Minutes IETF109: bess

Meeting Minutes BGP Enabled ServiceS (bess) WG
Date and time 2020-11-19 09:00
Title Minutes IETF109: bess
State Active
Other versions plain text
Last updated 2020-12-09

Working Group Status
 - RFC 7432 bis to be started as WG document. If any one has objection they
 should say. Its being presented
   first time in IETF109.
 - No RFC since last IETF
 - 4 draft in RFC queue
 - 7 drafts in AD review
 - fast-df-recovery - it would be ready before next IETF

draft-mishra-bess-ipv4nlri-ipv6nh-use-cases-01            Gyan Mishra      10
 - top 5 vendors have been contributing to this work for inter operability
 - presented in IETF108 as well. new author added since last IETF.
 - will request for adoption after IETF109
 - Acee : it is good as 5549 has been updated.
 - Gyan - As part of testing different vendor would test the NH encoding.

 - draft updated and implementation exist.
 - ready for WGLC

 - presented yhis draft since it will be making progress before next IETF.
 - adopted draft and implementation exist.

 - Wen :  how do we deal with broadcost traffic, clarification needed in draft.
          Inter subnet traffic wihh asymetric IRB , how is it going to get
   mankamana : will discuss in list / offline and add in draft.

 - based on implementation feedback modification is done.
 - some extra clarification added between Bridge Table, MAC VRF, IP VRF and
 other terms. - input needed from WG - draft will be published as WG document.

 - there are some area where author want to add more text . appendix to be added
 - getting ready for WGLC

Zeffry - instead of MP2P we can use ingress replication
Donald  - draft does use IR in draft.
Jorge -  do we send BFD discriminator per MAC route in BD.
Donald - more work needed , please provide suggestion for improvements.per PE
and per flow mechanism may be needed. Jorge - please include IP multicast , and
will send comment in list. Donald - will add it once get mail in list.

 - major change has been done in draft since last IETF based on WG input .
 - seeking more comment from WG

 - draft ready for LC
 - adopted long back
 - some typo fixing and some clarification added.

 - second time being presented in IETF
 - this version has cleanup
 - security considiration got added
 - adoption requested


Ali :  There is nothing new in this draft, all the issue with data plane
learning exist.  more detail to be sent over email to working group. all data
plane learning issue still exist in this draft Sami :  please send email to
list. ali - there would be flooding issue with all active multi-homing. Sami -
it may not be correct understanding. Tom hill: is it reinventing EVPN. does
this draft belong to BESS.  Does it belong to SPRING. since there is no BGP.
Sami: BGP part would be added in coming revision. Draft does not introduce or
use EVPN. We can discuss this in SPRING too. Tom hill : why do different
solution than EVPN Sami : trying to simplify solution then EVPN.

 - this draft provides information about use case for EVPN in mPLS
 - looking for WG comment, this has been discussed across multiple WG.
Stewart : it has been improved. how are you going to prevent re-assembly
lookup.  need to have discussion in MPLS group as well. may be propose using PW
headend. It is not clear what use case this is fixing jeffery: service provider
MTU can be different than customer. traffic source can not be asked to perform
the fragmentation. Acee: you might have to split signaling to transport. having
picture in draft would be good. some details in OSPFv3 needs to be changed.
Jeffery: pic would be used to describe it in more detail.