time.palemoon.org unreachable Topic is solved

[Stargate38]

If I try to sync with time.palemoon.org, I get "time synchronization failed". Could you please fix it? I checked, and the last time it successfully synchronized was April 30, 2026.

[athenian200]

Aren't there plenty of publicly available NTP servers, though?

Like I know for sure that time.windows.com is one, and also pool.ntp.org if I recall correctly. Any particular reason we would be running our own NTP server? I get not trusting publicly available services for some things, but accurate timekeeping seems like one of those things I'd take for granted that everyone would agree is needed...

[andyprough]

Any particular reason we would be running our own NTP server?

The discussion and the poll that led to this decision are here: https://forum.palemoon.org/viewtopic.php?f=17&t=32718

If I try to sync with time.palemoon.org, I get "time synchronization failed".

Yes, I'm getting "no eligible servers":

Code: Select all

$ ntpdate -q time.palemoon.org
ntpdig: no eligible servers

I'm thinking it was related to the palemoon.org server changes caused by VPS provider Afterburst shutting down, back in the April timeframe: https://forum.palemoon.org/viewtopic.php?f=17&t=33238

[Moonchild]

I'm not sure why ntpd wasn't running. I've restarted the service.

[andyprough]

I'm not sure why ntpd wasn't running. I've restarted the service.

Working now:

Code: Select all

$ ntpdate -q time.palemoon.org
2026-07-07 10:39:57.406732 (-0500) -0.001073 +/- 0.076584 time.palemoon.org 5.189.164.139 s3 no-leap

[Moonchild]

I investigated a bit further and it turned out the default ntpd configuration didn't have the service auto-restart after a panic threshold (bad served time). Not sure what the point is of having the default configuration be fragile like that, just to cater to the "convenience" of allowing a large time offset jump on first start (-g). A more sane default would have been not including -g, but have it auto-restart. If you're running the service daemon, after all, you're serving time to others (and would have configured multiple diverse upstream servers, mitigating any risk of serious time desync).
One of the time servers I had configured is no longer working, so that was probably the culprit. Now replaced with a stratum 2 of the National Research Council of Canada whom I trust will be reliable

[andyprough]

One of the time servers I had configured is no longer working, so that was probably the culprit. Now replaced with a stratum 2 of the National Research Council of Canada whom I trust will be reliable

I'm off by 0.005736 seconds. Which, coincidentally enough is the same amount of time that Mozilla executives are expected to work each week.

Code: Select all

$ chronyd -Q "server time.palemoon.org iburst"
2026-07-07T19:20:04Z chronyd version 4.6.1 starting (+CMDMON +NTP +REFCLOCK +RTC +PRIVDROP +SCFILTER +SIGND +ASYNCDNS +NTS +SECHASH +IPV6 -DEBUG)
2026-07-07T19:20:04Z Disabled control of system clock
2026-07-07T19:20:09Z System clock wrong by 0.005736 seconds (ignored)
2026-07-07T19:20:09Z chronyd exiting

[Veit Kannegieser]

also +1 recommendation for chrony from me, nice status display

Code: Select all

/etc/chrony>chronyc sources -v

  .-- Source mode  '^' = server, '=' = peer, '#' = local clock.
 / .- Source state '*' = current best, '+' = combined, '-' = not combined,
| /             'x' = may be in error, '~' = too variable, '?' = unusable.
||                                                 .- xxxx [ yyyy ] +/- zzzz
||      Reachability register (octal) -.           |  xxxx = adjusted offset,
||      Log2(Polling interval) --.      |          |  yyyy = measured offset,
||                                \     |          |  zzzz = estimated error.
||                                 |    |           \
MS Name/IP address         Stratum Poll Reach LastRx Last sample
===============================================================================
^* vps-fra1.orleans.ddnss.de     2  10   377   36m    -87us[ -123us] +/- 2176us
^- gw-001.oit.one                2  10   377   305  -5771us[-5771us] +/-   61ms
^- server2.as2.ch                2  10   377   374   +164us[ +164us] +/- 5998us
^+ cp.hypermediaa.de             2   9   377   369    +16us[  +16us] +/- 6642us
^- ptbtime2.ptb.de               1  10   377   475  +5570us[+5570us] +/-   10ms

[Moonchild]

Very similar to ntpq output which is what i use to check status of the ntp sync.

[Veit Kannegieser]

The similarity is intended, just that chrony shows different units ms/us. I had read claims that it tries to get better time accuracy.