[time] More Aggressive Server Prunning Redux
Tue Feb 26 10:08:46 UTC 2008
Dennis Hilberg, Jr. wrote on 26-2-2008 10:54:
> Jan Hoevers wrote:
> >> Interesting...
> >> That server is not open for ntpq queries, but it looks as if it's not
> >> syncing to anything except itself. What we see here is exactly the
> >> reason that we shouldn't use that LOCAL pseudo refclock driver.
>> Just noted that 188.8.131.52 is still at stratum 2. I's mis configured
>> no doubt, but I'm wondering how.
> 184.108.40.206 is actually synced to time-a.nist.gov and if you query
> time-a.nist.gov you get one source there, the local clock, although it is
> synced by the ACTS dial service. Maybe there's something wrong with the ACTS
> service and time-a.nist.gov is unable to sync, so it's wandering and
> 220.127.116.11 is following it blindly, apparently.
> If that's the case, you'd think that the ntpd at 18.104.22.168 would have
> declared time-a.nist.gov a falseticker a long time ago and chosen one of its
> other configured servers. Given that it hasn't though it might indicate
> there are no other configured servers, which is never a good configuration
> as we all can see.
No, not synced to time-a.nist.gov (why do you think so?), but to
22.214.171.124, which hasn't got a reverse dns and is reporting stratum 16.
Seems more like a bug than a misconfiguration. I'm giving up
understanding this. :)
More information about the pool