Path computation element: Difference between revisions

Revision as of 15:21, 23 June 2014

Revision as of 16:21, 10 December 2014

Line 10:

PCE represents a vision of networks that separates route computations from the signaling of end-to-end connections and from actual [[packet forwarding]]. There is a basic tutorial on PCE as presented at ISOCORE's MPLS2008 conference.<ref name=tutorial>Basic PCE tutorial [http://www.olddog.co.uk/Farrel_PCE_Tutorial.ppt "PCE Tutorial"]</ref>

Since the early days, the PCE architecture has evolved considerably to encompass more sophisticated concepts and allow application to more complicated network scenarios. This evolution includes Hierarchical PCE (H-PCE)<ref name=H-PCE>RFC 6805 [http://www.rfc-editor.org/rfc/rfc6805.txt "Hierarchical PCE"]</ref> and both Stateful and Active PCE.<ref name=Stateful>draft-ietf-pce-questionsRFC 7399[http://datatrackerwww.ietfrfc-editor.org/doc/draft-ietf-pce-questionsrfc/rfc7399.txt "Unanswered Questions in the Path Computation Element Architecture"]</ref>

A potential deployment of PCE separates the computation element from the client (the PCC) that request computation services. communications between the PCE and PCC are achieved using the Path Computation Element Communication Protocol (PCEP)<ref name=PCEP>RFC 5440 [http://www.rfc-editor.org/rfc/rfc5440.txt "Path Computation Element Communications Protocol"]</ref> which runs over the [[Transmission Control Protocol]] (TCP).