[ntp:questions] Timekeeping on Windows 2008r2 VM on Linux QEMU/KVM

Paul tik-tok at bodosom.net
Thu Jan 22 16:50:36 UTC 2015


On Thu, Jan 22, 2015 at 4:07 AM, Rob <nomail at example.com> wrote:

> The problem is that ntpd believes that corrections it is applying are
> because of frequency errors in the clock, while in this case they
> are because of resets done externally.
>
> During the startup phase, bad things happen anyway (like touching the
> carefully determined and saved drift value), doing a couple of successive
> restarts makes it worse because adjtime effects accumulate in the kernel
> without knowledge of ntpd.
>


I suspect the request was for more information not more speculation or
anecdotes.  It's probably okay to say "well it worked like that in 4.2.2"
but then the answer is update to a newer version.

Problems in the quoted message:
Confusing phase and frequency.
Proposing undocumented (mis)behavior regarding ntpd start-up (pre-SYNC).
Speculating that adjtime has memory.

And of course the OP (per the subject) is talking about Windows not Unix.

Applying a fixed offset outside of ntpd makes it worse.
>

I don't know what this means.


More information about the questions mailing list