You "quoting" is horrible. VEry hard to know what the old material and what your answer.
On 2012-06-23, ZHOU.Jun <dimotion.z...@gmail.com> wrote: > ??? 2012???6???22????????????UTC+8??????3???57???27??????David > Woolley????????? >> ZHOU.Jun wrote: > > > Thanks all for your feedback. > Why do you worry about small issues like a leapsecond when you tolerate > your customers to use software as old as that? First things first! > [ZHOU Jun]Since this old version was provided by us,we have to take the > responsibility. So why are you supplying such horribly old software? > > You know what? At the leapsecond time, everything will shake a bit > but within an hour or so everyone again agrees what the correct time > is and all servers are synced again. > [ZHOU Jun]I think I know, we have a bug in our software, we facing the risk > that kernel will be hanged when deal the leap second issue. How will the kernel hang? the system clock will keep ticking. ntpd will suddenly discover on its next poll that the local clock is out by a second. It will wait a while for a few polling intervals to make sure that this is not due to some temporary abberation, and then step the local clock (since it is out by more than 128ms. ) You could turn off stepping. > > seems rl sub-command is also work. _______________________________________________ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
