hi here are my comments on this document: Need to distinguish the different crankback methodologies when carrying out comparisons (source, per domain/area, etc.)Section 3 needs way better refined definitions before drawing conclusions - as examplest that section states:
"With the PD or PCE approach, the signaling of a TE LSP my fail because of lack of synchronization between the Traffic Engineering Database (TED) used by the computing entity and the actual link reservation state."
fair but why then the following is observed (Section 5.3)"Figures 2e and 3e shows the distribution of the total number of signaling failures experienced by the TE LSPs during setup. About 38% and 55% of TE LSPs for MESH-CORE and SYM-CORE, respectively, experience a signaling failures with per-domain path computation when link failures take place in the network. In contrast, only about 3% of the TE LSPs experience signaling failures with the PCE method. It should be noted that the signaling failures experienced with the PCE correspond only to the TEDs being out of date."
from"As discussed in the previous sections, signaling failures occur either due to an outdated TED or when a path cannot be found from theselected entry border router."
one can deduce that the delta is resulting from"With the PD approach, another situation leading to a signaling failure is when the selected exit domain router does not have any path obeying the set of constraints toward a downstream exit node or the TE LSP
destination."why the TED is not populated with that information for PD path computation purposes such to have comparable working assumptions ? (in other terms why only the BRPC method benefits from this advantage ?) comparisons shall be done with identical TED otherwise you don't need experiments by definition having less critical information leeds to such results
the observation is thus not that BRPC is better than PD computation but that BOTH techniques needs to have certain information about links inter-connecting domains to decrease the blocking/failure probability.
thanks, -d. Adrian Farrel wrote:
----- Original Message -----To: dimitri.papadimitriou@alcatel-lucent.be Cc: JP Vasseur <jvasseur@cisco.com>, Jaudelice Cavalcante de Oliveira <jau@cbis.ece.drexel.edu>, ccamp@ops.ietf.org From: Sukrit Dasgupta <sukrit@ece.drexel.edu>Subject: Comments/Feedback on draft-dasgupta-ccamp-path-comp-analysis-00.txtDate: Tue, 7 Aug 2007 11:54:20 -0400 Hi Dimitri, With respect to our discussion on the floor at CCAMP, it would be great to get your comments/feedback on draft-dasgupta-ccamp-path-comp- analysis-00.txt/pdf. This would help to further proceed with the draft. Thanks, Sukrit.