Just to start, this mailing list is for the discussion of updating OpenSolaris with NTP v4, not really for general NTP questions.
However, we are always glad to lend a hand. So, first, you should never in any version of Solaris or OpenSolaris, turn slewalways on without also specifying "disable pll". What made you think otherwise? NTP is designed to be aware that a leap second is coming. The protocol has flags that are sent to the client that tell it that there will be a leap second today. However, since leap seconds are not regular, there must be some human intervention somewhere to tell NTP that a leap second is coming. Hopefully, your NTP servers will do that. So assuming that the LEAP bits are set in the packets from the upstream servers, we have a few scenarios. First, when "disable pll" is specified and "slewalways" is also specified, then the clock will be slewed backwards starting at midnight. If "disable pll" is specified and "slewalways" is not specified, then the clock will be stepped backward. If "disable pll" is not specified, then the clock will step backwards. If the leap bits are not set, then xntpd will treat it like any other change in time from the servers. This means that it will take from 5 minutes to an hour after the servers all change before xntpd will notice and believe that the time has really changed. It will then step the clock (unless slewalways was specified) and this adjustment will cause the phase lock loop to skew the frequency adjustment. It will then be s few hours before the frequency adjustment gets back in place, during which time, the clock may or may oscillate a bit back and forth. tomohiro yamashita wrote: > In the case of Solaris 9, If we turn slewalways option on, xntpd will try to > always slew the time and will never make large steps backwards or forwards. > And also we have to set "disable pll". Because without this option xntpd will > not be able to correct time differences greater than 128ms. > > Now, When we turn slewalways option on, we do not have to set "disable pll". > Because xntpd can correct time differences greater than 128ms even without > this option. > > But, for a reap second insert event, I think we have to set "disable pll" > option. Without this option xntpd will make steps backwards even though > slewalways is on. No check of slewalways status at > http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/xntpd/xntpd/ntp_loopfilter.c#329. > > And, although slewalways option is on (with disable pll), step_systime() > function will not execute in the leap_process(). So, synchronization will not > happen just after leap second insert. I have to wait about 15min before start > synchronization. > > Why xntpd is implemented like that? -- blu "Murderous organizations have increased in size and scope; they are more daring, they are served by the most terrible weapons offered by modern science, and the world is nowadays threatened by new forces which, if recklessly unchained, may some day wreck universal destruction." - Arthur Griffith, 1898 ---------------------------------------------------------------------- Brian Utterback - Solaris RPE, Sun Microsystems, Inc. Ph:877-259-7345, Em:brian.utterback-at-ess-you-enn-dot-kom
