On Wed, Oct 24, 2018 at 05:55:08PM +0100, Joseph Billingsley wrote: > Ah just saw your reply to the other question from today. Is it the case > then that the only real option is a separate reference clock?
Yes. When measuring accuracy of a clock, a more accurate clock is needed as a reference. The same applies to measuring asymmetries in network. Two clocks are needed and they must not be synchronized over the network path on which we want to measure the asymmetry. > > On Wed, 24 Oct 2018, 17:47 Joseph Billingsley, <j...@josephbillingsley.co.uk> > wrote: > > > Thanks for your help! I have thought through the tests further and I think > > your first suggestion would work well for them. Do you have an opinion on > > other techniques for measuring the accuracy of the clock in a less direct > > way? Such as measuring the asymmetry in the master-slave and slave-master > > routes or measuring the offset from the master when a sync occurs. -- Miroslav Lichvar _______________________________________________ Linuxptp-users mailing list Linuxptp-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-users
