Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 IETF TEWG March 2003 Bandwidth Constraints Models for DS-TE Where do we stand? Francois Le Faucheur

Similar presentations


Presentation on theme: "1 IETF TEWG March 2003 Bandwidth Constraints Models for DS-TE Where do we stand? Francois Le Faucheur"— Presentation transcript:

1 1 IETF TEWG March 2003 Bandwidth Constraints Models for DS-TE Where do we stand? Francois Le Faucheur flefauch@cisco.com

2 2 IETF TEWG March 2003 Documented Models Russian Dolls Model (RDM) Maximum Allocation Model (MAM) Maximum Allocation with Reservation (MAR)

3 3 IETF TEWG March 2003 Concensus on RDM properties Where preemption can be used, RDM works very well: –simultaneously achieves Bw efficiency, isolation and protection against QoS degradation Where preemption can not be used, RDM works alright, but cannot achieve good isolation

4 4 IETF TEWG March 2003 Concensus on MAM properties MAM is "intuitive"/easy to conceptualise Where preemption can not be used, MAM is attractive: –it is good at achieving isolation –if one doesn't worry too much about "QoS degradation" of lower/medium classes, then MAM can achieve bw efficiency and isolation

5 5 IETF TEWG March 2003 Russian Dolls Model for DS-TE draft-ietf-tewg-diff-te-russian-01.txt F Le Faucheur : Cisco Systems, Editor Jim Boyle : PDNets,K Kompella : Juniper W Townsend : Tenor, D Skalecki : Nortel, T Nadeau: Cisco

6 6 IETF TEWG March 2003 Changes 00  01 Editorial alignment to –reqts-07: –WG decision to not have Default BC model –Editorial clean-up on usage of MUST Edited formula for Model definition Updated Security Considerations, as agreed on list

7 7 IETF TEWG March 2003 Open issues No remaining open issue on russian-01

8 8 IETF TEWG March 2003 Next Steps RDM is WG document and has been stable for several IETFs RDM properties well-understood (works very well with preemption, works reasonably well without premption) RDM represents a very attractive trade-off DSTE deployments in the field with RDM have started MAM can be used for specific environments where RDM is not suited  Proposal –Move diff-te-russian-01 to WG Last Call –Should this go Standards/Informational/Experimental ?

9 9 IETF TEWG March 2003 Maximum Allocation Model (MAM) for DS-TE draft-lefaucheur-diff-te-mam-00.txt F Le Faucheur : Cisco Systems Flefauch@cisco.com

10 10 IETF TEWG March 2003 MAM-related documents Requirements draft (diff-te-reqts): –Provides very brief definition of MAM –Not exhaustive analysis drafts (wlai-tewg-bcmodel, lefaucheur- tewg-russian): –Still do not provide exhaustive definition of MAM –Provide considerations on models properties –Cover other models than MAM –Somewhat biased (one for MAM, one for RDM)  We did not have a clean specification for MAM

11 11 IETF TEWG March 2003 -diff-te-mam Objectives: –provide clean MAM specification –“mirror-document” of diff-te-russian –Provide exhaustive (but brief) definition of MAM (incl when LOM is used) –Provide example CAC formulas –Only discuss MAM –Provide brief & non-controversial overview of MAM properties based on concensus view of the list –Content comes from other drafts and WG discussions  not expected to have any controversial content

12 12 IETF TEWG March 2003 Next Steps  Proposal –Accept -diff-te-mam- to WG document –Issue next rev to incorporate Waisum’s comments (Waisum will co-author) + add reference to BC Model analysis –Issue Last Call on next rev asap. –Progress BC Model analysis in a document separate from individual BC Model specs


Download ppt "1 IETF TEWG March 2003 Bandwidth Constraints Models for DS-TE Where do we stand? Francois Le Faucheur"

Similar presentations


Ads by Google