[Ntp] Re: NTP speed of synchronization question

[Miroslav Lichvar <[e-mail]>]

On Wed, Aug 14, 2024 at 10:09:33AM +0200, Toerless Eckert wrote:
> Assuming typical high-end routers and their implementation of NTP:
> 
> 1. Is there any reasonw why we should NOT be able to achieve synchronization of less than 5 msec ?

Not in a local wired network, but over internet routing asymmetries
can easily cause larger errors that 5 milliseconds. You would need to
limit the asymmetry.

> 2. Assuming synchronization only across some WAN connection, how long would it typically
>    take for a restarting router to synchronize against a peer to that level of accuracy or better ?

It depends on the NTP implementation and its configuration, most
importantly how frequently it is allowed to poll its server. It can
take a second or even less, or minutes, or hours.

> 3. Would synchronization to a local peer be faster ? If so, how fast could it be ?

It would be shorter by the round-trip time and the fewer number of
samples needed to get the required accuracy (less filtering need with
a smaller jitter).

> 
> 4. Am i correct to remember that the accuracy of synchronization should be accessible to
>    user of NTP (such as the aforementioned drafts mechanism) through some diagnostics interface ?
>    ( think to remmeber that to be the case from typical NTP implementation CLI).

The CLI tools should report the NTP root distance, or its two
parts called root delay and root dispersion. The distance estimates
the maximum error of the clock assuming the most extreme asymmetry in
the network delay.

-- 
Miroslav Lichvar