On 11/1/2010 11:54 AM, Jim Rowan wrote: > > On Nov 1, 2010, at 4:31 AM, Assarsson, Emil wrote: > >> Hi all, >> >> I would just like to have your input on this method of setting >> serverprefs based on the supernet distance. >> >> The background is that we have a large WAN with high latencies. The >> default method of calculating the weight for serverprefs is not good. >> We could have done this with static weighting but we need to keep >> things really flexible and self-configuring. > > There's a built-in assumption here that IP address assignments have a > clear relationship with latency. That's probably accurate in so far as > your company uses supernetting for routing. If you hand off Wan > connections to MPLS, for instance, that might no longer be a good > assumption, and this approach may fall down. This algorithm does appear > to be an improvement over the built-in one, though.
Jake Thebault-Spieker's GSOC 2009 project was to replace the built-in server preference algorithm based on classful addresses with one based upon the actual measured round trip times for the servers. This was implemented on Windows. The UNIX implementation was never finished but is sitting in gerrit. If someone wants to finish this work, then UNIX will have dynamic server preference updates as well. Jeffrey Altman
signature.asc
Description: OpenPGP digital signature
