[ntp:questions] NTP - orphan mode with SAME stratum ("tos orphan 6") FAIL to sync.

Martin Burnicki martin.burnicki at meinberg.de
Tue Apr 7 16:03:02 UTC 2009


Steve Kostecke wrote:
> On 2009-04-06, David Woolley <david at ex.djwhome.demon.co.uk.invalid> wrote:
>> Steve Kostecke wrote:
>>
>>> "Sync" is usually defined as "calculating the offset between a clock and
>>> a reference standard, stepping or initiating a slew to correct the
>>
>> I doubt that many would consider the initiation of a slew as being
>> synchronised.  I would suggest most people
> 
> s/most people/some people/

>From what I'm hearing from our customers, "most people" is the correct term.

>> would define synchronised as being when one achieves a high confidence
>> that the error will not subsequently exceed their required maximum
>> error, in the short term and will have a very low proportion of times
>> where it will in the long term.
> 
> It all depends on your application. One big problem in this forum is
> that too many participants assume that everyone is operating at the
> extremes.
> 
> If your application demands ultra high quality time you will protect
> your master clock(s) with an appropriate level of redundancy, high
> quality oscillators, back-up power systems, etc. Otherwise you'll have
> to wait until the offset is lost in the noise (however long that takes)
> after a restart.
> 
> Most users, though, consider their clocks to be "synced" when they see
> the asterisk indicating that ntpd has selected a sys_peer.

And I agree with David Woolley that most people expect ntpd to be
synchronized "as good as possible" if ntpd claims to be synchronized by the
leap bits set to 0, an asterisk diplayed in the output of "ntpq -p", and
the stratum following the stratum of its ref time source.

Martin
-- 
Martin Burnicki

Meinberg Funkuhren
Bad Pyrmont
Germany




More information about the questions mailing list