in clock_update (from
Dave Mills).
sys_mindisp is controlled by tos mindist and defaults to 1 millisecond:
[1] http://www.eecis.udel.edu/~mills/ntp/html/miscopt.html#tos
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
few hours when the 4.2.7p246 snapshot tarball is posted. You
can see a diff of last night's changes:
http://lists.ntp.org/pipermail/bk-ntp-dev-send/2012-January/002494.html
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
servers to spread the load.
For future reference, adding a server to pool.ntp.org is a
self-service operation via http://www.pool.ntp.org/ (don't omit
www.).
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org
?
That's a warning message from get_systime(). Dave Hart can say more,
and I think it means that a *very* slightly later call to get_systime()
managed to come up with a *very* slightly earlier time, based on the
random nature of the fuzz that we add below the precision floor.
That's basically
is retransmitted repeatedly until ACKed.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
On Wed, Dec 28, 2011 at 18:58, Rod Dorman r...@panix.com wrote:
Is this defined in an RFC or some other standards document?
http://standards.ieee.org/about/get/802/802.11.html
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
be slightly different, introducing apparent offset error.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
not replace servers
over time.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
UDP packets if they are lost rather than attempting to
retransmit them.
I do indeed, but UDP is treated as guaranteed by WiFi, and I expect
the reason is DNS over UDP otherwise becomes a user experience killer
due to extra seconds of wait for each loss.
Cheers,
Dave Hart
is ntpd assumes each remote address represents a single
oscillator. State is kept by ntpd clients for each association,
especially the clock filter register, so that unpredictable load
balancing will foul time transfer.
Cheers,
Dave Hart
___
questions mailing
for retransmission.
I wonder if the 90 minute periodicity to the -0.4 PPM shifts aligns
with some WiFi security renegotiation.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
. There are exceptions
where no high-resolution counter is used, such as often seen with ntpd
on Windows Vista and later. On such systems, reading the clock can be
as fast as reading 64 bits from shared memory, possibly twice if
partial updates are visible.
Cheers,
Dave Hart
enabled, you can consider the PPS offsets logged by the noselect
refclock a good measurement of the offset of the local clock.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
including LOCAL by default.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
hasn't come up
much. If you have any suggestions, please speak up here, or privately
if you prefer.
Thanks,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
a drift file and still calibrating
the drift.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
against 1.0.0.3 (1.0.0c) so I would hope that
would get along with the 1.0.0e DLL, but it may be worth trying the
older DLL.
Also I'm curious which part of the trace excerpt you posted shows a problem.
Cheers,
Dave Hart
___
questions mailing list
down to try to understand why its
responses are not lighting up the reach register.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
saw the problem even when building ntpd from source, this
seems unlikely to be the cause of your problem, but it seemed worth
mentioning. I'd suggest trying to determine what is causing the
bind() to ff02::101 to fail on your system.
Cheers,
Dave Hart
in 100ns units per 15.600 msec
fictional tick. The result is a unit adjustment to the
SetSystemTimeAdjustment argument results in a 6.4 PPM rate change.
sys_residual carry allows finer adjustments over multiple seconds.
Cheers,
Dave Hart
___
questions
to the present
value. Some ugly busy-looping logic might help refine that and also
overcome the incompatibility with newer Windows versions' clocks.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
and ASCII) is driven by the
length of the key value. If it is 20 characters or less, the ASCII is
used directly as the key value. If it is 21 characters or longer, it
is interpreted as hexadecimal-encoded.
Cheers,
Dave Hart
___
questions mailing list
). There is no kernel loop discipline
or kernel hardpps available on Windows, and the less-than-helpful
error message should be something like unimplemented.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo
? Are they still
supported in ntp 4.2.6?
I don't know what A, N, or S keys are, so I'll say no. But if you
tell me what they are, I might have a different answer for you.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
://www.eecis.udel.edu/~mills/ntp/html/authentic.html for an
example suitable for 4.2.6 and later.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
then
throwing some work at the machine increases the offset ntpd reports
for the NMEA association. If PPSAPI is working the timestamps are
snapped at serial port interrupt time and will not be affected by
additional ntpd scheduling latency.
Cheers,
Dave Hart
or ntpq -c rv 0 ctr_frequency.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
version suggests you might
have used documentation not matching your version. Steve Kostecke has
gathered and published a number of interesting versions of the NTP
docs at http://doc.ntp.org/ for easy reference.
Cheers,
Dave Hart
___
questions mailing
configure ntpd for remote management using symmetric key
authentication, you can use ntpdc's ifstats to retrieve per-address
statistics. In 4.2.7 ntpq also has ifstats.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org
On Mon, Nov 21, 2011 at 23:31, Pete Ashdown pashd...@xmission.com wrote:
Dave Hart daveh...@gmail.com writes:
PPS from the microseconds offset. Which raises my favorite question:
which version of ntpd? (ntpq -crv will say). I'm guessing it's 4.2.4
#ntpq -crv
associd=0 status=c618
of untrusted individual voices.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
On Thu, Nov 17, 2011 at 07:31, A C agcarver+...@acarver.net wrote:
On 11/16/2011 23:06, Dave Hart wrote:
I will probably try this later on another identical system.
Please do.
I'm going to
switch to a shared memory driver because my eventual desire was to have gpsd
running to collect
On Sat, Nov 12, 2011 at 05:52, unruh un...@invalid.ca wrote:
On 2011-11-12, Dave Hart h...@ntp.org wrote:
On Fri, Nov 11, 2011 at 20:23, A C agcarver+...@acarver.net wrote:
First attempt with gdb and a back trace after attaching gdb to the hung
process (note this particular running of ntpd
, please test and confirm for
yourself before relying on my expectation of success.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
the PPS, including cable delay and debouncing delay in
the UART before it recognizes the state change of CTS, and delay
between the UART asserting the interrupt and the interrupt routine's
execution, and the time to actually read the system clock.
Cheers,
Dave Hart
effects on mitigation which are
tricky to wrap your head around.
I encourage you to try a recent 4.2.7 ntpd with only the NMEA driver
with its PPS handling enabled
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org
On Thu, Nov 17, 2011 at 06:35, A C agcarver+...@acarver.net wrote:
On 11/16/2011 16:09, Dave Hart wrote:
You don't have to use two separate drivers for NMEA and PPS, even with
the signals coming in on different serial ports. At least in 4.2.7,
the NMEA driver tries /dev/gpsppsX for PPSAPI
On Thu, Nov 17, 2011 at 06:51, A C agcarver+...@acarver.net wrote:
On 11/16/2011 22:43, Dave Hart wrote:
On Thu, Nov 17, 2011 at 06:35, A Cagcarver+...@acarver.net wrote:
I compiled the latest development version 4.2.7p234. The NMEA driver
does not pick up /dev/gpspps for whatever reason
Ooops, see corrected suggested code:
On Thu, Nov 17, 2011 at 07:02, Dave Hart h...@ntp.org wrote:
change the if block to:
if (-1 == pps_fd) {
pps_fd = pp-io.fd;
msyslog(LOG_WARNING, %s cannot open %s: %d %m,
refnumtoa(peer-srcadr
On Tue, Nov 15, 2011 at 18:17, unruh un...@invalid.ca wrote:
On 2011-11-15, Dave Hart h...@ntp.org wrote:
The earliest http://doc.ntp.org/ snapshot referencing enable calibrate is:
http://doc.ntp.org/4.1.0/miscopt.htm
And the http://doc.ntp.org/4.1.0/refclock.htm under Driver Calibration
of the general idea than either
miscopt.html or refclock.html in the current documentation. 4.1.0's
diff (not tarball) was cut 1-Aug-2001.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
On Sat, Nov 12, 2011 at 03:19, Dave Hart h...@ntp.org wrote:
Excellent. I assume the stack trace is from ntpd 4.2.6p3. I think
you've found a bug in your system's libc dtoa() exposed by its
snprintf(s, %.2f, ...). I believe you will not be able to
reproduce the bug using 4.2.7
/book.html
[2] http://www.amazon.com/dp/1439814635
[3] http://tools.ietf.org/html/rfc5905
[4] http://www.eecis.udel.edu/~mills/ntp/html/warp.html
[5] http://www.pool.ntp.org/
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
);
cp += strlen(cp);
} while(i != start);
ctl_putdata(buffer, (unsigned)(cp - buffer), 0);
}
[1] http://www.jhweiss.de/software/snprintf.html
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
On Sat, Nov 12, 2011 at 05:52, unruh un...@chronyadvocacy.org wrote:
On 2011-11-12, Dave Hart h...@ntp.org wrote:
Excellent. I assume the stack trace is from ntpd 4.2.6p3. I think
you've found a bug in your system's libc dtoa() exposed by its
snprintf(s, %.2f, ...). I believe you
servers, too, when
they're self-organizing. Clients with relatively poorer clocks or
higher root dispersion will automatically be avoided as manycast
clients automatically cull servers that are not contributing for
extended periods.
Cheers,
Dave Hart
://www.eecis.udel.edu/~mills/ntp/html/rate.html#guard for an
explanation. This is the output or transmit counter described.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
myself than play pricing peek-a-boo, but then I'm doing this as a
hobby, not to solve a business problem. It's probably true that if I
have to ask the price, it will be higher than I'd be willing to pay.
Cheers,
Dave Hart
___
questions mailing list
of milliseconds.
Cheers,
Dave Hart
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
gdb without debug tracing
would allow enough poking around after the problem occurs to identify
it.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
ntpd at elevated
priority as the PPS timestamp is taken in the interrupt handler.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
a stack trace that hopefully goes back to
ntpdmain (that is, is hopefully on the correct ntpd thread, and not
any secondary or gdb-created thread).
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo
by the network stack at interrupt time or shortly thereafter,
rather than by ntpd querying the clock after select() returns.
When using a reference clock, or on a system without SO_TIMESTAMP such
as Windows, elevated priority improves the latency of the receive
timestamps.
Cheers,
Dave Hart
prefer.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
On Wed, Oct 26, 2011 at 09:32, Dave Hart h...@ntp.org wrote:
On Wed, Oct 26, 2011 at 04:07, Doug Calvert
dfc-l...@douglasfcalvert.net wrote:
Hello,
Why do the nmea and pps drivers differ regarding the default for flag2? How
would a user know which is appropriate?
PPS(22):
flag2 0 | 1
that goes into ntp's best estimate of the 'true' time,
which is irrelevant to this discussion)
Irrelevant if you want to paper over the minimum delay clock filter,
which you love to disparage and I view as a key error reduction step.
Cheers,
Dave Hart
/html/index.html
ntpdc.html is probably the best place to put advice about querying
mixed flocks. I appreciate your attention to this issue.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
/* 10 buffers initially */
#define RECV_LOWAT 3 /* when we're down to three buffers get more */
#define RECV_INC5 /* get 5 more at a time */
#define RECV_TOOMANY40 /* this is way too many buffers */
Good luck,
Dave Hart
,
please open a bug report at http://bugs.ntp.org./
Thanks,
Dave Hart
[1] http://tools.ietf.org/html/draft-odonoghue-ntpv4-control-01
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
, check your syslog for messages from
ntpd including the text time_pps.
Good luck,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
is, you might use a
gadget box to convert from TTL to proper 232 voltages. If that
gadget box inverts the DCD signal in the process, it's handy to be
able to select the clear edge as the on-time event.
Cheers,
Dave Hart
___
questions mailing list
questions
16000.0
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
enabled. Also nomodify is a subset of noquery, so doesn't need to be
listed with it in my example. See:
http://doc.ntp.org/4.2.6p2/accopt.html#restrict
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo
://bugs.ntp.org/2035 and please try with ntpq version
4.2.7p226 or later.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
such a
configuration, please compare peers vs. lpeers against a
broadcast/multicast client. How would you feel if peers were changed
to always show all associations as lpeers does?
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
2011/10/20 Miguel Gonçalves m...@miguelgoncalves.com:
2011/10/20 Dave Hart h...@ntp.org:
Better answer: Thanks to your report, as well as several others, I
found a bug that probably was responsible for mrulist not completing
for you. I also added progress indication, and the ability
selection.
Chers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
at default priority, you can safely point
the finger away from ntpd, IMO.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
example. See:
http://doc.ntp.org/4.2.6p2/accopt.html#restrict
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
the seconds. As a result, though, I'd never think to
use prefer in that configuration, so I don't have experience with NMEA
+ PPS while prefer is set on another association.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
2011/10/14 Miguel Gonçalves m...@miguelgoncalves.com:
Hi Dave!
2011/10/14 Dave Hart h...@ntp.org:
tock# ntpq -p
remote refid st t when poll reach delay offset
jitter
==
oGPS_NMEA(0
to be
able to similarly compensate its direct PPS synchronization.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
by the kernel
hardpps.
Once you verify things behave as expected without flag3 1, please file
a bug report at http://bugs.ntp.org/ noting the incompatibility of
flag3 1 and time1.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
exchanges with the server, where the delay is
determined with each exchange independently, insulating against
network changes.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
On Wed, Oct 12, 2011 at 08:03, Marco Marongiu brontoli...@gmail.com wrote:
Hi Dave and all
Il 11/10/2011 21:17, Dave Hart ha scritto:
I suspect not,
I had some IRC chat yesterday (my post was on hold for a while, and I
was too curious to wait :)
In short, mlichvar thinks I hit bug 1671
you're using.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
not, it would be nice to get it resolved before releasing a
new ntp-stable with all the 4.2.7 changes.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
chugging along, perhaps showing the number of
entries in the ntpq list.
If you use the server-side filtering (such as mincount=3) you're much
less likely to have any issues.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
to be prepared.)
Take care,
Dave Hart
P.S. As an aside, there are apparently some sample configurations
included by OSes in the wild which use restrict default kod without
the limited restriction. This is the same as not mentioning the kod
restriction, as far as RATE KoDs go, because without
for offering NTP,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
go on a diet and lose one event code, or we
need to steal a status word bit from another field to expand the event
field.
Thanks for reporting the contradiction. Please consider filing a bug
report at http://bugs.ntp.org/
Cheers,
Dave Hart
___
questions
,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
, a reported tick of 15.600 msec (156000) isn't unheard
of, but I can't point to a system using that value where ntpd behaves
well, either.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
2011/9/27 Miguel Gonçalves m...@miguelgoncalves.com:
2011/9/27 Dave Hart h...@ntp.org
I think you're simply confused. The FreeBSD sysctl you modified isn't
adjusting the TSC frequency. It's adjusting FreeBSD's estimate of the
TSC frequency. I stand by my understanding that a positive
you narrow down the failure,
though.
Good luck,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
2011/9/26 Miguel Gonçalves m...@miguelgoncalves.com:
2011/9/26 Dave Hart h...@ntp.org
2011/9/25 Miguel Gonçalves m...@miguelgoncalves.com:
I have a machine running FreeBSD 7.4 and it seems that the CPU clock
runs
too fast or slow (positive offset in loopstats is fast or slow?).
ntpd's
as the binary key material. If it
is 21 characters or more, up to 40 characters are decoded from hex
resulting in up to 20 bytes of binary key material. No matter the
encoding, no more than 20 bytes of key material are used.
Cheers,
Dave Hart
___
questions
that period, the frequency compensation is locked
while ntpd attempts to slew away the initial offset until less than
500 usec remains, or the 300s limit elapses.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org
2011/9/25 Dave Hart h...@ntp.org:
2011/9/25 Miguel Gonçalves m...@miguelgoncalves.com:
tick# tail -10 /var/log/ntp/loopstats
55829 66814.311 0.10871 185.398 0.03540 0.002317 4
55829 66831.312 0.10025 185.398 0.00984 0.002167 4
55829 66847.312 0.11242 185.398 0.01303
to do so, or using the more
sophisticated google cosine-shaped smear to limit the rate of change
of the leapsec-induced frequency adjustment. The slewing approach has
the upside of no risky backward step, and the downside of up to 1s
offset vs. UTC during the slew.
Cheers,
Dave Hart
. The disadvantage, particularly with such a short-lived
smear, is that interval timing that starts or ends during the special
two seconds will be inaccurate by up to a second.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http
On Mon, Sep 19, 2011 at 10:55, Samuel Nyarko sknnyy2...@gmail.com wrote:
On Sep 15, 6:30 pm, Dave Hart h...@ntp.org wrote:
Please try commenting out all restrict lines in ntp.conf and repeat
your test. There's a good chance that will fix things, and you can
iterate over restrict changes
is ignored because historically the ntpd Windows
port's priority-raising code failed to take it into account, and
always attempted.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
, however.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
with a buggy ntpd.
Cheers,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
and good luck,
Dave Hart
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
);
return (0);
}
As the comment notes, not all PPSAPI implementations can actually
timestamp either edge. So it's not a limitation of ntpd, but of
NetBSD's timepps.h, in your case.
Cheers,
Dave Hart
___
questions mailing list
questions
201 - 300 of 639 matches
Mail list logo