[ntp:questions] NTP sync failing with error: Leap not in sync
stenn at ntp.org
Thu Apr 2 19:29:36 UTC 2015
Jochen Bern writes:
> On 04/01/2015 10:23 PM, Unknown wrote:
> > 10.0.0.5: Server dropped: Leap not in sync
> > stratum 3, precision -23, leap 11, trust 000
> > 1 Apr 14:28:05 ntpdate: no server suitable for synchronization
> > found
> That's indeed an error code labelling your server as unsuitable as a
> sync source to your clients. I admit I've never seen it together with
> another stratum than 15 (also an "is unsynced" marker), though.
S15 is still sync'd and valid - it just isn't useful to serve time to
anybody else because 16 is "unsync'd" so when you add 1 to the stratum
you are unsync'd at S16.
> The stratum being given as 3 suggests that your server considers itself
> synced to one of its stratum 2 sources, and my first guess would be that
> that server signals weird flags as well. (Which *should* cause your
> server *not* to consider it suitable, either.)
> Running ntpq against your server doesn't output its peers' leap seconds.
> Try dry-running ntpdate against them from your server.
Better yet, run the latest sntp against it. We decode and display the LI
bits now, so you can also easily see the leap bits.
> > 2 packets reassembled into response
> NTP packets as used in syncing shouldn't be large enough to trigger any
> MTU problems short of downright insane pMTUs. Do you have *duplicated*
> packets happening in your network?
> > server 0.centos.pool.ntp.org iburst
> > server 1.centos.pool.ntp.org iburst
> > server 2.centos.pool.ntp.org iburst
Better to use:
pool 0.centos.pool.ntp.org iburst
Harlan Stenn <stenn at ntp.org>
http://networktimefoundation.org - be a member!
More information about the questions