| Re: [chrony-users] NTPsec's NTPViz Grapher Log Adapter |
[ Thread Index |
Date Index
| More chrony.tuxfamily.org/chrony-users Archives
]
- To: chrony-users@xxxxxxxxxxxxxxxxxxxx, Miroslav Lichvar <mlichvar@xxxxxxxxxx>
- Subject: Re: [chrony-users] NTPsec's NTPViz Grapher Log Adapter
- From: Matt Corallo <ntp-lists@xxxxxxxxxxxxxxx>
- Date: Thu, 17 Sep 2020 11:10:42 -0400
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mattcorallo.com; s=1600353689; t=1600355442; bh=UbwBZMhHrA1HTmqn2bz/2I0c+wuOxbf9NH2tGHZepVg=; h=To:References:From:Subject:Date:In-Reply-To:From; b=BnNOUfZNjsudSmALJ0GqFUomo5zejOt767RksN40W5cw7/IRy4i0aO9/DzVI+kj6I 9sOUTKTG06Mx+/XxIZPu2+JSDYzF3LrBZ2A60J8emKR+U7W8WDRsuwfogfzhZZ/WuD kBU/GiYiaIqLkDIlcOxpj/nWS5Zixe9hcZaAB3RrOtwGeYJfCHVbPRYq3KHGWsHec/ kJ6DGUkVHvBGRf4eQkfDJRX8Dy61oN96Ng9qaolSaHpXNILsgvE/LodEQHAg9Pqijy 58aouDz/vDPN2G7eEB8Qi4ZyJ/XnGtaH1VimD6lnjjxKdDOW1bTABvo23MYMZ/An4H 2jkxk6UKb4OXg==
On 9/17/20 2:58 AM, Miroslav Lichvar wrote:
On Wed, Sep 16, 2020 at 04:45:03PM -0400, Matt Corallo wrote:
In case anyone else finds this useful, I spent the time to write out some
python to convert chrony's stats logs to something that looks close enough
to ntpd that ntpsec's ntpviz can graph it.
Interesting. If you turn it into a project, I can add a link to the
"Useful links" page. IIRC ntpviz was originally based on the
chrony-graph project, so this seems to make it full circle :).
Ha, fun. I put a readme on it at https://github.com/TheBlueMatt/chrony_ntp_logconv (which probably makes some false claims).
Its not perfect - "Local RMS Frequency Jitter" is just the error bound
chrony reports (which seems to be something similar, but different?)
Different, but both are measures of the frequency stability.
and
statistics.log doesn't report an average/median RTT time, so that is filled
with 0s (maybe that would be a useful thing to provide?), but it otherwise
works.
Measured delays are reported in the measurements log. In the
statistics log there is a number of currently held samples, so with
some processing of the history you could find a median, minimum, or
whatever you like.
Right. Maybe when I find the time. Its noted as a todo.
Demo at https://noc.as397444.net/ntpgraphs/ script at
The ~30ms offset shift happening for one of the GPS refclocks every
~1.5 days looks odd. Any idea what is causing that?
That's 1.5 hours, not days, but, yea, no idea, that's the serial port and noselect'd, just there for reference. Ask
u-blox why their UBX output jumps on the 9-series, I guess? I didn't bother to check if its a gpsd bug/quirk, though.
Matt
--
To unsubscribe email chrony-users-request@xxxxxxxxxxxxxxxxxxxx
with "unsubscribe" in the subject.
For help email chrony-users-request@xxxxxxxxxxxxxxxxxxxx
with "help" in the subject.
Trouble? Email listmaster@xxxxxxxxxxxxxxxxxxxx.