1 / 10

draft- akiya - mpls -entropy- lsp -ping

draft- akiya - mpls -entropy- lsp -ping. IETF 88, Vancouver, Canada. Nobo Akiya George Swallow Carlos Pignataro Nagendra Kumar. Background. RFC4379 D efined an object for label load-balancing Did not define usable procedures RFC6790 “Updated” RFC4379 But not with a complete description

ardara
Download Presentation

draft- akiya - mpls -entropy- lsp -ping

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. draft-akiya-mpls-entropy-lsp-ping IETF 88, Vancouver, Canada Nobo AkiyaGeorge SwallowCarlos PignataroNagendra Kumar

  2. Background • RFC4379 • Defined an object for label load-balancing • Did not define usable procedures • RFC6790 “Updated” RFC4379 • But not with a complete description • In particular no way defined to put an entropy label in the Label Stack • Doesn’t cover PW or FAT PW cases • Doesn’t cover ELI/EL imposed along the path • Doesn’t cover hardware that can’t handle ELI/EL • Technically not an update at all by IETF procedures

  3. Entropy Label FEC Stack Entry • ELI does not need a new FEC, it can simply go in the NIL FEC • Defined a new sub-TLV for EL • Value of label carried here • NIL FEC is not explicit enough in the PW, FAT PW cases

  4. FAT PW • Procedure for FAT PW is fairly simple • Sender • Bottom of FEC Stack ::=EL • Next FEC pushed is PW FEC • Transport FEC(s) pushed • Mid-point • Values of LB Type 9 successively replace value in EL

  5. PW • A bit less clean • Trace using EL FEC instead of PW FEC • Advantage works exactly like FAT PW • Value of PW label goes into bottom of FEC Stack TLV and the actual FEC stack of the packet • Disadvantage – need to ping destination twice • Once to know you have arrived • Second time to verify PW • Alternative is to define new PW FEC types for when you have control words

  6. LSP Ping/Trace on Entropy Motivation • In general, 4 load balancer types • IP based LB, not imposing Entropy • IP based LB, imposing Entropy • Label based LB, not imposing Entropy • Label based LB, imposing Entropy • Restore ECMP trace capability when mixed load balance techniques in • One LSP • Stitched LSPs • Clarification on Multipath Type {9}

  7. LSP Ping/Trace on Entropy • Echo request • Multipath info to carry IP address set, Label set or both IP address and Label set. • Echo reply • Communicate LB technique(s) to initiator • And one of below in multipath info: • IP address subset • IP address subset and corresponding entropy labels • Label subset • Label subset and corresponding entropy labels • None due to lack of multipath info match

  8. LSP Ping/Trace on Entropy Extensions • DS Flags: • L: 1=Label based LB, 0=Non-label based LB (i.e. IP) • E: 1=Imposes EL, 0=Does not impose EL • Multipath Information Type {10} • IP Multipath Information • Label Multipath Information • Associated Label Multipath Information • FEC definition for Entropy label • Procedures on how to use above for initiator and responder nodes

  9. Next Steps • Work in progress • The more we delved into it the more we found • Would like the WG to review – more eyes, perspectives will help

  10. Backup Slide(in case we need to discuss with topology) 1.1.1.1 1.1.1.2 1.1.1.3 A B C 1.1.1.5 1.1.1.6 D E F 1.1.1.4 H G I 1.1.1.7 1.1.1.8 1.1.1.9

More Related