[Ntp] Antw: Re: Antw: Re: Calls for Adoption -- NTP Extension Field drafts -- Four separate drafts

["Ulrich Windl" <[e-mail]>]

>>> Miroslav Lichvar <[e-mail]> schrieb am 12.09.2019 um 11:46 in
Nachricht <20190912094603.GN21704@localhost>:
> On Mon, Sep 02, 2019 at 05:52:33AM ‑0700, Harlan Stenn wrote:
>> The thing is that NTP is a complete specification with a goal of time
>> synchronization that includes *known behavior in the face of impulse
>> response*.
>> 
>> If one uses different algorithms with different impulse response
>> behaviors, this can trigger oscillations and even unstable tracking (the
>> wheels fly off the bus).
> 
> This was discussed a bit on the last meeting.
> 
> Does anyone have an example of a PLL that is unstable when cascaded
> with a different loop? If something works well enough to be stable

I think it was the old Dave Mills consensus like: "It's only NTP if it
provides the same output for the same input". So if you think of some
improvements, they still would not be allowed (I had some discussions while
implementing PPS in the Linux kernel)

> when synchronizing a clock of an NTP client to a stable reference
> clock in different network conditions, how could it cause its own
> clients to become unstable?
> 
> ‑‑ 
> Miroslav Lichvar
> 
> _______________________________________________
> ntp mailing list
> [e-mail] 
> https://www.ietf.org/mailman/listinfo/ntp