Hello! According to http://www.eecis.udel.edu/~mills/ntp/html/prefer.html when selecting new system peer on the final stage the following happens:
"...The survivor list is then sorted by increasing root distance and the first entry temporarily designated the system peer. ..." Imagine there's ntp server and reference clock. Reference clock always has root distance equal to zero, so it will always be prefered to ntp ( unless ntp has "prefer" keyword ) So, reference clock could have any dispersion less then 16000 to be selected. For example, imagine that GPS receiver just appeared from nowhere, has only one element in filter and dispersion of about 7900. So for me it's a very bad clock which properties are almost unknown to us. According to the algorithm such clock would be preferred to any ntpd server. Please, could someone explain why this was implemented in this way? May be it will be wiser to make a selection based on dispersion, not a "root distance" ? Thanks, Nickolay _______________________________________________ questions mailing list [email protected] http://lists.ntp.org/listinfo/questions
