Rob Janssen writes: > Harlan Stenn wrote: > > Rob Janssen writes: > >> Harlan Stenn wrote: > >>> Miroslav Lichvar writes: > >>>> On Thu, Nov 06, 2014 at 06:08:32AM -0500, Jared Mauch wrote: > >>>>> One thing I will highlight here is we have ips in both v4 and v6 > >>>>> pools that are anycast and there may not be a good way for something > >>>>> to realize 4 ips may just be 2 hosts with the same reference. > >>>> So the problem is that a client could be unknowingly using the same > >>>> server twice, which would effectively give it two votes in the source > >>>> selection? Can this happen with addresses from one DNS reply or only > >>>> over time when some sources are replaced? > >>> We probably want to take another look at what a "refid" means as for as > >>> loop detection goes. > >> I don't expect loops to be a problem. > > If a pool server has a machine with both an IPv4 and an IPv6 reference, > > what's to stop that server from being selected twice, once for each IP? > > > > Or am I missing something? > > That is not a loop. You might consider it an unwanted situation, but > it is not dangerous like a loop would be. > > The issue of IPv6 refids has been discussed for over a decade, but > apparently there still has been no real solution, only kludges.
It will cause that machine to be "considered" more strongly, which is not right. When I'm awake next I'll consider the case where A has both v4 and v6 IPs, and B talks to A via v4, while C talks to A via v6, and then see what happens when other machines talk to B and C (and other combos). And again, I'm real tired and haven't thought clearly about this. H _______________________________________________ pool mailing list [email protected] http://lists.ntp.org/listinfo/pool
