1 / 16

CDNI FCI Analysis

CDNI FCI Analysis. CDNI Working Group IETF 89 London March 2014 presented by Daryl Malas (D.Malas@cablelabs.com) on behalf of Matt Caulfield (mcaulfie@cisco.com). Background. Two major competing proposals for the CDNI Footprint and Capabilities Interface (FCI)

lola
Download Presentation

CDNI FCI Analysis

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. CDNI FCI Analysis CDNI Working Group IETF 89 London March 2014 presented by Daryl Malas (D.Malas@cablelabs.com) on behalf of Matt Caulfield (mcaulfie@cisco.com)

  2. Background Two major competing proposals for the CDNI Footprint and Capabilities Interface (FCI) Volunteered at Vancouver meeting to produce a detailed analysis comparing the proposals Presentation summarizes email from Feb 22

  3. draft-seedorf-cdni-request-routing-alto-06 Primarily based on the ALTO protocol • Each dCDN hosts an ALTO server, uCDN is ALTO client • ALTO Network Map indicates “footprint” (PID) • Property Map provides capabilities per footprint • Future extensions cited for incremental updates

  4. draft-ma-cdni-capabilities-04 Primarily based on CDNI-specific representation • Each dCDN hosts an HTTP server, uCDN is HTTP client • Each capability has a name, list of values, and an optional list of footprints • Each footprint has a type, list of values, and a mode • Incremental updates via HTTP POST with seq #s

  5. Transport & Encoding draft-seedorf • HTTP transport • JSON encoding draft-ma • HTTP transport • JSON encoding JSON chosen by CDNI WG as default encoding

  6. Data Representation draft-seedorf • ALTO Network Map and Property Map • Leverages existing work done by ALTO WG draft-ma • CDNI-specific syntax for footprints and capabilities • Custom solution for CDNI Fundamental difference and decision point for WG

  7. Hierarchy draft-seedorf • Footprints have capabilities • More intuitive in the cascaded CDN case (footprints from many dCDNs may be concatenated) draft-ma • Capabilities have footprints • Less intuitive in cascaded CDN case(requires merge of multiple capabilities)

  8. Cost Information draft-seedorf • Loosely described using ALTO Cost Maps draft-ma • No solution described Importance unclear, given business rules may override

  9. Extensibility & Versioning draft-seedorf • Not described draft-ma • Not described Details lacking, a clear gap in both drafts

  10. Dependencies draft-seedorf • Depends on multiple Internet Drafts from ALTO WG • Leverages existing error handling, security, etc. draft-ma • CDNI development and consensus Fundamental difference and decision point for WG

  11. Capability Inheritance draft-seedorf • PID Property Map rules for implicit inheritance may add complexity to implementations draft-ma • Completely explicit capabilities (no inheritance) Explicit capabilities is a better approach

  12. Update Notifications draft-seedorf • No method described for receiving update notices draft-ma • Asynchronous HTTP POST from dCDN to uCDN • May violate RESTful principles POST is a concrete approach, but potentially flawed

  13. Incremental Updates draft-seedorf • Uses ALTO Incremental Update proposal • Relies on JSON Patch for encoding draft-ma • HTTP POST header indicates seq# of update • Footprints include mode for overwrite vs append Both approaches reasonable, but new HTTP header may not be achievable

  14. Conclusions • Both drafts well-written and good starting points • WG must decide if the benefits of reusing ALTO syntax and semantics outweigh the costs • Benefits: existing error handling, security, encoding, scale, proposed ALTO drafts for extending functionality • Drawbacks: dependence on ALTO WG, some inflexibility • Recommend we focus on a simple HTTP GET before attempting to solve incremental updates (if ever)

  15. Recommendation

  16. A New Hope • Focus on developing a common FCI object • Base the syntax on the FCI semantics draft • Jan & Kevin will team up, probably starting with Kevin’s object as baseline • More help welcome! • Look beyond address prefixes to other footprint indications (like region), focus on capabilities • Do the work as a CDNI deliverable • Simultaneously, work on transporting this FCI object • Define a new ALTO service for this purpose • Would probably be done in the ALTO WG (by whom? You!) • We’ll talk about that in ALTO (provided this goes well here…) • Potentially consider alternative transports here if necessary

More Related