Richard, Just addressing the FRR comments.
Another point is that IETF does not create standards based solely on research projects. For a whole new routing protocol to be considered, implementation and some deployment experience is generally needed. IRTF may be interested in considering routing research. Curtis In message <of7b611f3a.6f9bf7df-on48257cd9.0018f932-48257cd9.001c3...@zte.com.cn> [email protected] writes: > > --------"Network convergence doesn't follow link state > dynamics - Fast reroute exists. " > > Comparison of convergence time: The settings of OSPF spf_delay and > spf_hold_time can affect the change of convergence time. The > convergence time of the network with 2480 nodes is about 15-20 > seconds(as seen in the following pages); while the FAR does not need > to calculate the SFP, so there is no such convergence time. These > issuesstill existin rapid convergence technology of OSPF and ISIS > (such as I-SPF). The convergence speed and network scale constraint > each other. FAR does not have the above problems, and the convergence > time is almost negligible. > > And test data is been include in another pptx material named OSPF in > DCN (2).pptx, which can be download from IETF. > > Looking forward to further discussion. > > Best. > > Richard Bin Liu Your comments above seem to ignore the point about fast reroute (FRR) stated above "Network convergence doesn't follow link state dynamics - Fast reroute exists", though the comment could have been better worded. MPLS RSVP-TE FRR and IP/LDP FRR are both order(1) or at worst scale with route install speed. No SPF is required for FRR to provide protection, restoring traffic flow though sometimes on a suboptimal path. The goal of FRR is to keep protection time under 45 msec (and old and some argue meaningless SONET target). In practice many FRR implementations provide protection in under 20 msec even for large topologies. The convergence which follows only serves to further optimize traffic flow, particularly where RSVP-TE is used. One commercial CSPF measured about a decade ago completed in 30-40 msec on a test topology of 450. That was on a 300 MHz or less PPC or Pentium-2. Todays processors are an order of magnitude faster, so we could expect (with order NlogN scaling of SPF) to get about the same SPF time on a topology of 4K or more nodes with no improvements in software. The CSPF was in an RSVP-TE implementation with a lot of ancilliary data structures built for later incremental CSPF and for caching of CSPF results. I don't remember the plain ol' SPF results but I think it was under 10 msec and wasn't an issue so wasn't studied to the same extent. Curtis _______________________________________________ rtgwg mailing list [email protected] https://www.ietf.org/mailman/listinfo/rtgwg
