Dear Martin, Thanks for your answer (Readers of the ntp:questions list did not reply my more deatiled question http://lists.ntp.org/pipermail/questions/2015-May/040388.html )
> > > Only config I have vs leap is same as in 2012: > > > > > > leapfile /var/db/ntp/leap-seconds.list > > > > Yes, all newer NTP servers of us is configured such a way and > > they predict the next leap second well. > > However in this case the procedure is bit more complicated. > > ntpd 4.2.6p5 should work fine with a proper, current leap second file. Ooops! I speak about 4.2.2p1. Sorry for confusion. There is no problem with 4.2.6p5 servers. > What irritates me is that in your original email you wrote: > > > However if you are interested in it, I can tell that the best result > > with the latest leap-seconds.3629404800 file and 4.2.6p5 is this: > > > > $ ntpq -c "rv 0 leap,tai,leapsec,expire" one_of_my_servers > > leap=00, expire=201505281348, leapsec=201501050000, tai=35 > > $ > > > > The expire value is the minute of restarting ntpd, > > and the leapsec was at 5th of January. That is strange... > > There have been faulty 3rd party GPS receivers which started to announce a > leap second shortly after the satellites started to broadcast the leap second > information for June 30. For me this sounds like some software tool has > generated the leap second file you are using to pass that announcement to > ntpd. Actually _all_ stratum1 upstream servers are managed by me and all of them is configured correctly: $ ntpq -c "rv 0 version,leap,tai,leapsec,expire" an_upstream_server leap=00, version="ntpd [email protected] Wed Feb 4 09:27:04 UTC 2015 (1)", tai=35, leapsec=201507010000, expire=201512280000 $ Meanwhile $ ntpq -c "rv 0 version,leap,tai,leapsec,expire" old_strata2_server leap=01, version="ntpd [email protected] Thu Nov 26 11:34:34 UTC 2009 (1)", expire=201506010846, leapsec=201501050000, tai=35 $ looks bad. > Can you send your leap second file to me and/or the list? Do you have a clue > where that file comes from? I downloaded this file: ftp://time.nist.gov/pub/leap-seconds.3629404800 md5sum is d1a4caab3153fbeea4d6230282cef117. Gabor _______________________________________________ pool mailing list [email protected] http://lists.ntp.org/listinfo/pool
