ntpMerlin ntpMerlin v3.x

[MvW]

Oh, but I had no intention of causin' the boss extra work. He has other thing on his head now anyway. I was just curious about the default drift value. And I'm here to learn so I feel free to ask.

As for the temperature, I recently did some research (well, actually, I DuckDuckGo'd a bit around) about the fluctuations I see and came to the conclusion that temperature is appparently assumed the biggest factor in these deviations. Check for yourself when the space where your router is (in my case, living room, on a high cabinet, next to a central heating element, but actively cooled by a 120 mm Noctua 12V fan on about 5V) is at it lowest temp and what happens when the next morning the room gets actively heated again. As these deviations are a natural occurence and they just are what they are, I see no reason to ask the boss to find a way to compensate them. The router detects and reacts to environmental temperature changes and the air blown through it and we can see it for example in our charts. That's all. Right?

 

[Wade Coxon]

, see no reason to ask the boss to find a way to compensate them.

Thats the thing though. The fact that you have a drift value is a direct result of the chrony daemon actually compensating for the drift already.

When chrony fetches time from the upstream servers, it is able to detect when the local clock is consistently drifting out of synchronisation. It measures the drift, and from then on will consistently apply an anti-drift factor to the local clock in between updates from the upstream servers.

There is a reason that ntpd was the only game in town for a long time. Timekeeping is a very subtle and clever trick .

 

[Wade Coxon]

Now I think about it, what might be nice to see on the GUI page is the ability to query the source information from the time daemon ("chronyc sources" or "ntpq -p" depending on the daemon selected).
Maybe just a text box that can be updated on demand via button-click would do. No need to schedule a regular update I don't think.

 

[zaxcom]

My drift tends to run around 15 ppm. I see most others posts with much lower amounts. Does this signify anything about my router?

 

[BreakingDad]

To use the words of a good friend... it's working bloody amazing! On RT-AX88U, RT-AX86U, RT-AC86U, RT-AC3100, RT-AC68U, RT-AC66U_B1. All running RMerlin Beta 3 or better.

Is your friend either british, aussie or south african?

 

[heysoundude]

Is your friend either british, aussie or south african?

If he was paraphrasing me, you're close. Commonwealth close. I am Canadian.

 

[Wade Coxon]

My drift tends to run around 15 ppm. I see most others posts with much lower amounts. Does this signify anything about my router?

Nothing to worry about. Just natural variation. Mine runs between 10 and 11ppm.

My old server PC used to run at about 30ppm from memory.

 

[raion969]

heyy on my rt-ac5300 router i cant get the webui on ntp merlin going? does anybody have the same problems ?

 

[MvW]

Only for ntpMerlin only or are you unable to access the WebUi at all? Have you just upgraded?

Otherwise login with ssh and execute

service restart_httpd

if you're unable to access the webUI.

 

[DanNuke86]

Updated list:

pool time.cloudflare.com iburst nts # Anycast
server netmon2.dcs1.biz iburst nts # Singapore
server ntp1.glypnod.com iburst nts # San Fransisco
server ntp2.glypnod.com  iburst nts # London
server ntpmon.dcs1.biz iburst nts # Singapore
server nts.netnod.se iburst nts # Sweden
server nts.ntp.se iburst nts # Sweden
server nts.sth1.ntp.se iburst nts # Sweden
server nts.sth2.ntp.se iburst nts # Sweden
server nts.time.nl iburst nts # The Netherlands
server ptbnts2.ptb.de iburst nts # Germany
server ptbnts3.ptb.de iburst nts # Germany
server ptbtime1.ptb.de iburst nts  # Primary server for Europe
! server ptbtime2.ptb.de iburst nts (currently not operational) # Germany
! server ptbtime3.ptb.de iburst nts (not yet operational) # Germany
! server timemaster.evangineer.net iburst nts (The Netherlands, times out for unknown reasons)

I've added a comment with the location of the server. Not sure whether a comment behind a line could cause trouble in chronyd.conf, so if you seen any errors, please remove the comment including the hashtag.

! server ptbtime2.ptb.de iburst nts (currently not operational) # Germany
! server ptbtime3.ptb.de iburst nts (not yet operational) # Germany

These 2 servers are now operational.

ptbtime1.ptb.de              NTS     1   15  256   93    0    0    8  100
ptbtime2.ptb.de              NTS     1   15  256   92    0    0    8  100
ptbtime3.ptb.de              NTS     1   15  256   93    0    0    8  100

 

[MvW]

! server ptbtime2.ptb.de iburst nts (currently not operational) # Germany
! server ptbtime3.ptb.de iburst nts (not yet operational) # Germany

These 2 servers are now operational.

ptbtime1.ptb.de              NTS     1   15  256   93    0    0    8  100
ptbtime2.ptb.de              NTS     1   15  256   92    0    0    8  100
ptbtime3.ptb.de              NTS     1   15  256   93    0    0    8  100

I heard Friday it was a matter of days now. Good to hear their both online. Thanks for the update. I’ll edit the list tomorrow.

 

[XIII]

We should use ptbtime[123].ptb.de in the configuration file, not ptbnts[123].ptb.de, right?

Zeitsynchronisation von Rechnern mit Hilfe des "Network Time Protocol" (NTP) - PTB.de

Die Physikalisch-Technische Bundesanstalt (PTB), Braunschweig und Berlin, ist das nationale Metrologie-Institut mit wissenschaftlich-technischen Dienstleistungsaufgaben. Sie misst mit höchster Genauigkeit und Zuverlässigkeit – Metrologie als Kernkompetenz.

www.ptb.de

 

[MvW]

We should use ptbtime[123].ptb.de in the configuration file, not ptbnts[123].ptb.de, right?

Zeitsynchronisation von Rechnern mit Hilfe des "Network Time Protocol" (NTP) - PTB.de

Die Physikalisch-Technische Bundesanstalt (PTB), Braunschweig und Berlin, ist das nationale Metrologie-Institut mit wissenschaftlich-technischen Dienstleistungsaufgaben. Sie misst mit höchster Genauigkeit und Zuverlässigkeit – Metrologie als Kernkompetenz.

www.ptb.de

From Martin Langer, creator of NTS:

ptbnts1.ptb.de is now the official NTS-secured time server of germany and central europe.

Source: https://weberblog.net/setting-up-nts-secured-ntp-with-ntpsec/

Ultimately it's up to you to decide which ones you wish to use in your config. You add them both (both ranges), chronyd will decide for itself which it prefers.

The list has been updated.

 

[MvW]

My connections to ptbnts[1/2/3].ptb.de keep timing out, so I'll try to find out why and get back to you. If these are supposed to be main server(s) for Germany and Europe, you'd expect they'd be reachable and I was informed that no separate key/certificate should be required anymore.

 

[MvW]

My connections to ptbnts[1/2/3].ptb.de keep timing out, so I'll try to find out why and get back to you. If these are supposed to be main server(s) for Germany and Europe, you'd expect they'd be reachable and I was informed that no separate key/certificate should be required anymore.

I contacted Physikalisch-Technische Bundesanstalt to find out why ptbnts[123].ptb.de are timing out. If and when I get a reaction I'll let you know.

 

[MvW]

I can't find it in the documentation and it's about the only output of chronyc where it's not specified, but am I correct that the field Reference Time in the output of

chronyc ntpdata

is displayed in UTC?

 

[pdc]

My drift tends to run around 15 ppm. I see most others posts with much lower amounts. Does this signify anything about my router?

The router is built with a "good enough" crystal, but not that critical since it is assumed you are going to have some form of time sync. 15 ppm is still less than a minute per month, so as a wall clock you might not even notice.

That said, drift is the the rate by which the system’s clock would be wrong if chronyd was not correcting it. On my router with chronyd, the corrected rate (residual frequency in chronyc tracking) is a teeny 0.001 ppm (1 part per billion!) or one second every thirty years!

So run some form of regular time sync and it doesn't matter what the uncorrected drift would be since it is the job of time sync to compensate for drift.

 

[EmeraldDeer]

I can't find it in the documentation and it's about the only output of chronyc where it's not specified, but am I correct that the field Reference Time in the output of

chronyc ntpdata

is displayed in UTC?

I doubt anyone here would know for sure. One can infer from the description "The NTP values from the last valid response." that it is UTC because NTP syncs globally to UTC. Inputs of other formats are converted to UTC. Outputs to local timezones are converted by the operating system.

 

[heysoundude]

a poll rate of 2^6 seconds (poll 6) from your router to any server or pool will be plenty- that's a ping every 64 sec. they'll barely notice. and they'd rather your router do it than every device on your network.
I wouldn't go lower - thats when you become the mosquito in the dark room...except they see your IP ringing their doorbell too often, so tread carefully and lightly

 

[MvW]

I contacted Physikalisch-Technische Bundesanstalt to find out why ptbnts[123].ptb.de are timing out. If and when I get a reaction I'll let you know.

I had an early reply this morning from the NTP administrator at PTB, stating that ptbnts[123].ptb.de were only active for a limited time, for testing purposes, and that now that ptbtime[123].ptb.de are fully operational since last weekend, the ptbnts[123].ptb.de servers have been taken offline. I will remove them of the list. It seems that ptbtime1.ptb.de is now leading for Germany and Europe and fully NTS compatible, as well as #2 and #3, but I'll ask for confirmation on that.

Dear Mr. [redacted],

thank you for your interest in our NTS servers. I am sorry you couldn't contact the three test servers. Recently we extended the service on our public NTP servers so now they also run NTS. The test servers are not available anymore. For more information see https://www.ptb.de/cms/en/ptb/facha...ters-using-the-network-time-protocol-ntp.html.
If there are further questions feel free to contact me.


Kind regards
[redacted]

Physikalisch-Technische Bundesanstalt
The National Metrology Institute of Germany
Q.42 - Zeitverteilung mittels IP / Time Distribution via IP
Bundesallee 100
38116 Braunschweig
Tel.: +49 [redacted]
Fax: +49 [redacted]
E-Mail: ntp-admin@ptb.de

The list has been updated: http://www.snbforums.com/threads/ntpmerlin-v3-x.68508/post-680984