Re: Using iburst and pools

[Gary E. Miller]

Yo Steve!

On Thu, 19 Mar 2020 11:30:35 -0500 (CDT)
Steve Bourland <address@hidden> wrote:

> Yo Gary!
>    Hope you are well during this chaos, reading the mailing list
> brought up a couple questions:
> 
> >  And I've complained about the iburst before.  As well as using the
> >   
> pool.
> 
> Can you (easier than me Googling it) point me to the discussion?  I
> think I understand the iburst, but what's bad about using pool
> servers?

Oh, I see the confusion.  I have no issue with iburst and pool.  Usually.

To back up a tad, iburst sends an initial flurry of time requests to the
chimer.  This primes caches from here to China a back, fills the DSP/PLL
loops with data, and through various mechanisms, gets ntpd closer to
"correct" time quicker than normal on startup.

Note the detail: quicker on startup.  Otherwise no effect.

The problem at hand is that a user has local PPS and wants that time
to win the startup race, ahead of the more jumpy network chimmers.  It
is maybe 1,000x more accurate and stable.

To that end, a user places the PPS at the top of the ntp.conf, with a
short poll time than other refclocks, so that ntpd will select it first.

But, when you add in a bunch of marginal pool servers, from all over
the planet, and iburst them, then they win the race instead of the nice
local PPS.  ntpd starts out with the worst refclocks, not the best.

Eventually ntpd will sort it out, usually, but best to just pick the
local PPS on startup.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        address@hidden  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin

pgpmBpVl5tKGI.pgp
Description: OpenPGP digital signature