On 8/14/08 6:32 PM, Brian Carpenter allegedly wrote: > [re-sent because once again psg.com is rejecting messages from gmail] > > On 2008-08-14 18:49, Flinck, Hannu (NSN - FI/Espoo) wrote: >> Hello Brian >> >> Can you bit elaborate what do mean by multipath management? It seems to >> go beyond <src, dst, sport, dport> tuple. Having multipath management in >> the hosts requires some visibility into the network topology beyond >> having multiple src, dest addresses or multiple locators. > > Not necessarily, if you believe Mark Handley. In his model, transport > congestion management automatically finds the best paths without any > topological knowledge. And in shim6, the reachability detection finds > working paths without any topoloogical knowledge. In these cases > we don't need any routing-type imformation in the hosts. > > Brian
The scheme Mark presented (leaving out all the other multipath ideas being tossed about) can specify the endpoint addresses but not the paths taken to reach them. It will in theory find which address pairs experience the least congestion, but there is no reason to think that paths to different endpoint addresses will be different until you get close to the end site. I believe that explicit TE in the core is still going to be useful. Scott -- to unsubscribe send a message to [EMAIL PROTECTED] with the word 'unsubscribe' in a single line as the message text body. archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg
