[questions] Re: iburst and burst options in NTP

[David Woolley]

On 02/06/2023 18:17, Mohammed Siddiqi wrote:

I am currently working on making sure NTP works according to the RFC 5905 and 
we use NTP.org implementation of ntpd. I have been going through the RFC and 
NTP.org’s documentation


If you want a definitive implementation of the RFC, you need the ntp.org 
version that was current at the time of its publication, and represents 
the reference implementation.  The current source code will include 
improvements.


I suspect Dave Mills considered the RFC to be more like a research 
paper, describing things as they were at the point at which it was 
written, rather than as a standard, frozen for all time.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscr

[questions] Re: new ntp server organization

[David Woolley]

On 11/04/2023 21:28, David Woolley wrote:

On 11/04/2023 18:21, Jakob Bohm wrote:

I wonder when and why this advice changed from 3 to 4.  When I started


I'm pretty sure that Byzantine General protection has been advised for 
more than 15 years.  It's in the NTPv4 RFC, so it is at least just short 
of 13 years. <https://www.ietf.org/rfc/rfc5905.txt>


I meant to include this quote:

/*
 * There must be at least NSANE survivors to satisfy the
 * correctness assertions.  Ordinarily, the Byzantine criteria
 * require four survivors, but for the demonstration here, one
 * is acceptable.
 */
--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: new ntp server organization

[David Woolley]

On 11/04/2023 18:21, Jakob Bohm wrote:

I wonder when and why this advice changed from 3 to 4.  When I started


I'm pretty sure that Byzantine General protection has been advised for 
more than 15 years.  It's in the NTPv4 RFC, so it is at least just short 
of 13 years. 

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubsc

[questions] Re: new ntp server organization

[David Woolley]

On 07/04/2023 10:47, Renzo Marengo wrote:

I know NTP server numbers must be 1, 3 or 4,


I don't understand this.  NTP servers don't have numbers.  They do have 
a stratum, but that is determined automatically from the current time 
distribution tree.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Lebanon DST (was: Iran dst)

[David Woolley]

On 23/03/2023 14:45, David Woolley wrote:

Microsoft also prefer a lead time of at least a year, for such changes


It would appear that Lebanon has made a change with less than 3 days 
notice.  They didn't inform IANA, so again they had to rely on private 
individuals.  The IANA change (in 2023b) was made with just over a day 
to spare, but it will probably take weeks for OS maintainers to issue 
the update.  Microsoft may well be unaware, but even if made aware, it 
is very unlikely that they will issue an update before it is too late.


(In fact there seems to be an internal dispute in the country as to 
whether the change has happened, see 
<https://www.bbc.co.uk/news/world-middle-east-65079574>.)


Again this is not an NTP problem, as NTP times are always  UTC and 
converted by the OS.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Iran dst

[David Woolley]

On 22/03/2023 17:37, David Woolley wrote:
I wonder if this could be the result of an embargo, implemented by 
either Iran on the USA?  At least for open source software, the latter 
should not be an issue, as it can be obtained from other countries.


Looking at 
<https://techcommunity.microsoft.com/t5/daylight-saving-time-time-zone/bg-p/DSTBlog> 
it would appear that Microsoft haven't acknowledged the existence of the 
change, which may explain the problem with PCs, as Microsoft operating 
systems reportedly dominate.


I think this may, in part, be because they are not permitted to provide 
the Windows Update service to Iran, although I found it difficult to 
find the exact current position with respect to that.  They may also be 
ignoring official Iranian government communications, or the government 
may not have tried to communicate (the IANA update seems to have been 
triggered by a private individual).


It seems to me that the government could have anticipated this and 
provided guidance (e.g. to disable DST, or even registry patches to 
manually correct the information - although not as easy to modify as the 
IANA data, it looks to me that the Dynamic DST registry entries are 
fairly straightforward).


Microsoft also prefer a lead time of at least a year, for such changes, 
although I believe there would still have been interim change notices on 
the above blog page, if they had accepted the existence of the change.


I haven't found a similar page for Apple.

In case the point is getting lost, none of this relates to NTP, which, 
like GPS, etc., does not have  time zones.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
U

[questions] Re: Iran dst

[David Woolley]

On 22/03/2023 15:53, Y J wrote:

and mobile phones


My Samsung Android has the correct time zone information (as checked in 
the world clock tab of the clock app).  It is possible that the local 
mobile networks have the wrong information and that telling Android 
phones to use their actual location, rather than the network supplied 
information, might help.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Iran dst

[David Woolley]

On 22/03/2023 17:37, David Woolley wrote:

and so has the IANA master copy of the database


This change was committed on the Github version of the database on May 
10, 2022 
<https://github.com/eggert/tz/commit/66b18d9835dda089d65349c86e7851c4f0746e09>


It looks like the change was published in version 2022g of the database, 
on 2022-11-29.


This is the citation for the change:

# From Ali Mirjamali (2022-05-10):
# Official IR News Agency announcement: irna.ir/xjJ3TT
# ...
# Highlights: DST will be cancelled for the next Iranian year 1402
# (i.e 2023-March-21) and forthcoming years.


(It is conceivable that the change date being so near the the Iranian 
year end confused some companies not using the TZ database, but all open 
source software, should be using this database which seems correct.)

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Iran dst

[David Woolley]

On 22/03/2023 16:59, Jim Pennino wrote:

DST is NOT controlled by ntp, but rather your operating system's time
zone file.


Furthermore, Debian Linux has the correct rules:

TZ=Asia/Tehran date; TZ=UTC date
Wed 22 Mar 20:48:32 +0330 2023
Wed 22 Mar 17:18:32 UTC 2023

and so has the IANA master copy of the database 
 (file: 
asia):


RuleIran20172019-   Mar 21  24:00   1:00-
RuleIran20172019-   Sep 21  24:00   0   -
RuleIran2020only-   Mar 20  24:00   1:00-
RuleIran2020only-   Sep 20  24:00   0   -
RuleIran20212022-   Mar 21  24:00   1:00-
RuleIran20212022-   Sep 21  24:00   0   -

# Zone  NAMESTDOFF  RULES   FORMAT  [UNTIL]
ZoneAsia/Tehran 3:25:44 -   LMT 1916
3:25:44 -   TMT 1935 Jun 13 # Tehran Mean Time
3:30Iran+0330/+0430 1977 Oct 20 24:00
4:00Iran+04/+05 1979
3:30Iran+0330/+0430

This says the current standard offset from UTC was three and a half 
hours, since 1979, the last change was to no additional offset from 
that, and was at midnight on September 21 2021, and no future changes 
are known.


I wonder if this could be the result of an embargo, implemented by 
either Iran on the USA?  At least for open source software, the latter 
should not be an issue, as it can be obtained from other countries.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Does anyone know why would server return Key id: 0?

[David Woolley]

On 01/03/2023 10:10, Mihovil Pupovac wrote:

 Key id: 0


A crypto NAK: 
--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Is it possible to connect ntp server via other port?

[David Woolley]

On 01/01/2023 07:24, Jacob Bingham (UFTS) wrote:

I would also like to know the answer to this question, as port forwarding port 
1234 to port 123 is problematic for me and I just want to use ntpdate -q 
localhost (but on port 1234)


Looking at the documentation, you can change the source port from 
ntpdate, but you cannot change the port on the server.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Specify IP for outgoing sync

[David Woolley]

On 28/09/2022 23:13, Blažej Krajňák wrote:


I read the documentation a few times but can not find if it's possible to 
spcify IP for outgoing connections?


NTP uses UDP, which is connectionless.


I have server with multiple IP addresses and I need to specify the concrete one 
to be used to sync with servers/peers. For now, ntpd selects one of the 
listening interfaces (still the same one, but I can not determine why this one).


I would imagine you could do it at the OS level, by ensuring you have 
(static) routes to the remote address via specific interfaces.  I can't 
think why ntpd would bind to anything other than 0.0.0.0.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questi

Re: [questions] GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 21/06/2022 01:31, chris wrote:

Sorry David, but that is incorrect. If you look at the app note you
posted, input capture section, quite clearly states that:

 > When the rising or falling edge is detected at the CCP1 pin,
 > the interrupt flag CCP1IF bit is set.



If you go to a more primary source 
, you 
will see, in the diagram on page 15-3, that the logic path for raising 
interrupts diverges from that for doing the capture, after the event 
detection, so the capture is not dependent on the interrupt occurring.


Also, on page 15-4, that the actual interrupt can be selectively disabled.

In my original response, I did meant to add that there would typically 
be an interrupt, but that the interrupt could be handled with low 
priority (you will get a good result any time until the timer actually 
wraps round.  Even then, the use of an interrupt is not a fundamental 
requirement.


For a dedicated appliance, interrupts can actually be a liability, as 
they can introduce race conditions.  Nonetheless, I think all current 
microcontrollers support them.


With precision time, over Ethernet, one would expect the normal Ethernet 
device driver interrupt handling, but, for example, the device driver 
might well service several incoming messages on one interrupt.  However 
capturing the event time still means that nano-second accuracy is possible.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: NTP community feels broken

[David Woolley]

On 21/06/2022 10:48, William Unruh wrote:

Who cares if it is no longer RS232, if it works. As has been pointed out
virtually no rs232 hardware is RS232 according to the standards.


Oddly that is the point I was trying to make.  Someone was insisting 
that you mustn't cut corners, and must use an RS232 driver when 
interfacing to a nominally RS232 input port.  I was analyzing the 
consequences of actually conforming to RS232.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 20/06/2022 21:32, David Woolley wrote:
Actually I think something like that is pretty standard for most 
microcontrollers.


E.g. see <https://www.electronicwings.com/pic/pic18f4550-timer-capture>
--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 20/06/2022 15:54, chris wrote:

Still doesn't explain how the pps state change gets into the
system at hardware level. Devil in the detail, as usual.
It sounds like a polled system, not an interrupt driven one,


Interrupt system are I think pretty much always polled at the hardware 
level.



which can never be as accurate as the latter. Assuming a poll
rate of say, 1mS, not unreasonable for a modern cpu, Then if
the pps state change is not seen at one poll, but is seen at
the next, then the uncertainty can never be better than that


The uncertainty is the clock period in the capture hardware, because the 
hardware counts and returns the the number of its clock cycles, so you 
know very precisely how long it was from the the PPS to when you polled. 
 Actually I think something like that is pretty standard for most 
microcontrollers.


I believe there are Ethernet cards that do this to support, I think, 
Precision Time Protocol, which I suspect is what high frequency traders use.



1mS, since it cannot be known when during that 1 mS the state
transition occurred. Best results for ntp really need a true
real time os, but it's  far more common to use general purpose
os's because of the out of the box functionality., but not
ideal.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: NTP community feels broken

[David Woolley]

On 20/06/2022 00:14, chris wrote:

Early processors
often had just a couple of interrupt pins, irq and nmi,


You mean microprocessors.  NMI seems to be a concept that was introduced 
with microprocessors, and multi-priority interrupts existed before even 
the earliest microprocessors.


In terms of software, I don't think typical Linux code really fully 
supports multi-level interrupts, and, if anything, may effectively 
invert the priority order.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 20/06/2022 00:32, chris wrote:

Don't really understand what's meant by "input capture timer" ?. Is
that related to some polling method, or what ?...


I believe they mean a hardware clock, that can be read directly by the 
software, and is also latched into a register when the PPS signal 
arrives.  A variation might be to start a hardware timer when the PPS 
arrives, again making it directly readable.


Also, whilst NMEA is low priority, I suspect it is not uncommon to have 
its generation scheduled at the PPS time, making it much closer to the 
last PPS than the next.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 19/06/2022 16:17, Jim Pennino wrote:

which is why I would question your "which should be a DB-25".


It's RS232's should be.  Actually, Wikipedia seems to say that the D 
version says must be.  A lot of this thread is about RS232 compliance, 
and part of that compliance is using the correct connector.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: NTP community feels broken

[David Woolley]

On 19/06/2022 16:10, William Unruh wrote:

So teminate it to get rid of the refections.


It's no longer RS232 if you terminate it anything less than, ISTR, 4k, 
or reverse terminate in anything other than 300 ohms.


I think 100 ohm is typical for the characteristic impedance of a wire 
pair, so given the open source voltage of the driver is 12 volts, by 
terminating it you have removed all your noise margin, if you accept the 
transition region is the full +/-3V.


If you want to use an EIA standard that uses terminated lines, you 
should use RS 422, which also uses differential pairs, which also 
improves noise immunity.   Or RS 423, which, although it uses a 
differential line receiver, is unbalanced, with one of the pair tied to 
local functional ground at the transmitter.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 19/06/2022 13:46, Jim Pennino wrote:

CTS/RTS is pin 8 on a RS-232 connector, so how is that "PPS over USB"?


CTS is on pin 5, and RTS on pin 4, on an RS232 connector, which should 
be a DB-25 one.


The DE-9 connector, used on PCs, is a TIA-574 connector, not an RS232 
one, and RTS is on pin 7 (although not useful in this context, as it is 
output from the PC).

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: NTP community feels broken

[David Woolley]

On 19/06/2022 11:30, chris wrote:


Yes, it does have a spec for the slew rate, but i'm not sure that would
always be met for modern devices, as such uarts can often be run at data
rates of up to 230 Kbytes / second, or about a 4uS bit time. That
implies much faster rise and fall times, which relates to slew rate.


But those aren't RS232 compliant devices, although the slew rate limit 
is still low enough to support those data rates, but not with the full 
cable length.


We can theorise about that, but iirc, I bought two ttl to rs232
converters at the time. If I can find the second one, i'll set it up
on the bench and measure the propagation delay on a scope from input to
output. Also, see what effect the 2.5nF cap has on the waveform and 
timing...


The 2.5nF is mainly intended to represent the capacitance of the cable, 
which is deliberately operated with both source and load termination 
impedances well above the characteristic impedances, so behaves like 
capacitor.  With the full length cable, rather than the lumped 
capacitance, you would expect the voltage to staircase upwards, each 
time the leading edge reflection returns.


The 2.5nF basically allows for the 50 foot maximum cable at 50pF/ft.
--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: NTP community feels broken

[David Woolley]

On 19/06/2022 01:06, chris wrote:

In practice, that will be small, since the
data sheet figures for a typical max232 assume a 2.5nf capacitive
load on the output, whereas a few inches of wire into a rs232 line
receiver setup might be much faster.


As we are talking about compliant RS232, which is the only real world 
reason for not just connecting TTL directly to the RS232 port, the the 
2.5nF condition is the maximum capacitance that can appear across the 
driver output as the result of what it is driving.  That sets a minimum 
possible slew rate.


However a compliant RS232 system also has a maximum permitted slew rate, 
intended to minimise cross talk, and probably also to ensure that long 
cables don't ring, as the initial transient reflects backwards and 
forwards.  30V/µs is the maximum permitted slew rate for a compliant 
system.  If your system exceeds that, even if it is using RS232 drivers, 
it is not a compliant system.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 19/06/2022 07:38, Daniel O'Connor wrote:

OK, then to which of the USB connector pins do you connect the PPS
signal to get "PPS over USB"?



You can connect them to CTS or RTS, on FreeBSD these can then hook into the 
kernel PPS API.

It works very well in practise, especially for the cost & effort required.


Taking the other side for this one, which of Ground, D+, D-, and VBus 
have alternative names of CTS and RTS? (I guess I should add StdA_SSRX-, 
StdA_SSRX+, StdA_SSTX+, and StdA_SSTX-, for USB 3.0, compatibility.)

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 17/06/2022 20:45, Jim Pennino wrote:

Have fun writting the necessary device driver...


You can buy chips preloaded with the relevant code for the encode side, 
for single figure sums and most OSes already include the decode side code.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 17/06/2022 20:16, Terje Mathisen wrote:
The key idea is of course that in order to know where a GPS is located 
with better than 3 m precision, the unit by implication also knows what 
time it is, to within 10 ns of UTC(USNO). The only problem is to be able 
to convey that info to a connected NTP server.


It only needs the time relative to the visible constellation, as the 
ultimate position accuracy depends on time differences not absolute 
time.  Absolute time errors affect position at about 7.5mm/µs.


 quotes ≤30 
nanoseconds, 95% of the time.  That probably includes uplink allowances. 
 Also, I'm not sure if UTC() exists in real time.  UTC proper only 
exists some weeks after the event.  There might have to be an allowance 
for the difference between the instantaneous estimate of UTC(USNO) and 
the final value.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 17/06/2022 15:01, Jim Pennino wrote:

OK, then to which of the USB connector pins do you connect the PPS
signal to get "PPS over USB"?


D+ and D-, using for example a Communications Device Class module to 
encode it for transmission.  I guess HID would be more appropriate, for 
an isolated digital signal, but HID often uses low rates.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 17/06/2022 16:34, chris wrote:

As for compatibility, while a mismatched connection may work, it's bad
practice to do that, where you are dealing with microsecond timing
and want to avoid jitter. Use the correct interfaces and do the job
right, then you can fit and forget:-)...


RS232 isn't optimal for PPS as it is slew rate limited to 30V/µs, which 
means it will take at least 0.5µs from a resting level until it has 
fully cleared the transition region, if implemented to standard.


TTL is the more natural logic family for high accuracy PPS.

GPS should be able of achieving time transfer accuracies of better than 
.03µs.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: NTP community feels broken

[David Woolley]

On 17/06/2022 19:14, Paul G wrote:

Where is it in this 
tarball:http://www.eecis.udel.edu/~ntp/ntp_spool/ntp4/ntp-4.2/ntp-4.2.8p15.tar.gz

If it's not there then you're probably in the wrong list/group.


This group is about the NTP protocol, not just the version 4 reference 
implementation.  chrony, the SNTP client that Debian use, etc., are also 
on topic.  I would say that the management of the pool servers was 
definitely on topic.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 17/06/2022 00:55, chris wrote:

No argument with that, but some have tried to bypass a converter,
feeding the ttl pps into the rs232 port, which may work in some
cases. TLL pps low level, in particular, won't guarantee the rs232
input line to switch, whereas, of course, the ttl high will switch.
The rs232 ip needs zero or a minus level to properly work and avoid
jitter...



In practice, RS232 line receivers will emulate the characteristics of 
the 1489 or 1489A ICs (which are still available).  In the simplest 
configuration for those, the input low threshold voltage is no less than 
+0.75 volts and the input high threshold voltage is no more than +2.25 
volts, which makes nearly all real life true RS232 line receivers TTL 
compatible.  (The A variant has a larger hysteresis, and results in a 
higher high threshold.)


The main reason for not using TTL are that is isn't designed for driving 
more than a few inches of wire, and it doesn't have the huge noise 
margins of true RS232 drivers.


In terms of level converting to TTL, the decades old 1488 line drivers 
are also still available, although they need +/- 12V supplies, as are 
the newer, but still decades old, MAX232 devices, that have charge pumps 
to derive the 12 volts from a 5 volt supply.  As such, there is no 
really sensible reason to re-purpose device intended for shifting 
between 3.3 an 5 volt CMOS levels.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: NTP community feels broken

[David Woolley]

On 17/06/2022 00:30, chris wrote:

lack of full disclosure, documented


I'm having trouble understanding what this means.  If you mean that the 
documentation is poor, that is a common problem with open source 
software, as it relies  on volunteer effort, and programmers don't like 
writing documentation.


Actually, in terms of end user documentation, I find most technology 
poor.  Businesses tend to document the small number of marketing claims, 
in marketing language, and don't provide detailed functional descriptions.


For software, the only really good documentation is often the test plan, 
but that is considered highly confidential for commercial software. 
Open source coders are less likely to write formal test suites.


The original author of ntpd saw it as a maths problem, and was 
frustrated by the inability of the RFC system to cope with mathematical 
notation.  The primary documentation, the version 4 RFC, is written as a 
maths paper, but with the Greek letters spelled out.


ntpd will have been documented to the same sort of detail as an 
experimental rig in an academic paper.  The main detail would have gone 
into the particualr property of the core algorithm that the paper was 
trying to investigate.


It may also be significant that its primary developer is now 84.

Although I say lack of documentation is a problem for all technology, 
what I sometimes find out with hardware is that it is all based on a 
small number of special purpose ICs, and if you can establish what is 
being used you can get a long way towards a real specification, rather 
than the half page marketing hype on Amazon, by looking at the 100 page 
data sheet for the ICs.  Semiconductors seem to be an area where 
detailed end user documentation is still available in the public domain.


It is common for the consumer products to be more or less direct 
implementations of the typical application circuits, from the IC data 
sheet.  This doesn't work so well with software, as every user can end 
up customising its use to a level that only the final manufacturer would 
do for hardware, and they are more prepared to do so than the people who 
sell products on Amazon.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 16/06/2022 15:54, Thibaut HUMBERT wrote:

I have a serial port, but I don't know how to convert the PPS output (0 / 3.3V) 
to RS232 (-5V / +5V).


RS232 is +/-12V, although, input values of +/-3V are unequivocal.  In 
practice line receivers have both positive and negative going thresholds 
> 0V, and various things rely on this to put lines into the right state 
when the plug is removed (false for control lines and marking(?) for data).

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: GPS+PPS vs NTP server, why a huge offset ?

[David Woolley]

On 16/06/2022 09:01, Thibaut HUMBERT wrote:

When I modify the PPS pulse length in u-center, the offset varies:


I would suggest you are detecting the wrong edge of the pulse.  You may 
need to add an inverter.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Date and time are wrong.

[David Woolley]

On 10/06/2022 08:47, joshua wrote:


connect to the right timezone


NTP doesn't use timezones.  All servers use times based on UTC.  Any 
timezone issue is a local problem, in your operating system.

--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

[questions] Re: Please Document the Peer Command (and Let the World Know How to Use It)

[David Woolley]

I think you meant pool, not peer, in the subject...[more interleaved]

On 26/05/2022 22:22, Frank Wayne wrote:


It is mentioned on the Access Control Commands and Options page 
(https://www.eecis.udel.edu/~mills/ntp/html/accopt.html#restrict), but there's no reason 
someone would think to look for it there. Someone trying to use pool for the first time 
with "restrict default nopeer" is presented with a .PEER. refid in ntpq, but is 
offered no clue as to why no peers actually show up.


If they hadn't looked there, they shouldn't have include a restrict 
command in the first place.




The first article 
(https://www.digitalocean.com/community/tutorials/how-to-configure-ntp-for-use-in-the-ntp-pool-project-on-ubuntu-16-04)
 is actually how to stop ntp from using pool. The author literally instructs 
the user to remove the pool commands in the default ntp.conf and replace them 
with server commands.


Which is quite correct in the context of the article, which is about 
becoming a member of the pool, not a user.  You don't want an incestuous 
situation where pool members are getting their time from the pool.




Before I move on, it is interesting to note that -- as this "documentation" describes --  
Ubuntu's official repository provides a default ntp.conf that contains the following pool commands 
(that the "documentation" instructs us to remove):

pool 0.ubuntu.pool.ntp.org iburst
pool 1.ubuntu.pool.ntp.org iburst
pool 2.ubuntu.pool.ntp.org iburst
pool 3.ubuntu.pool.ntp.org iburst
pool ntp.ubuntu.com

This would, I think, produce as many as twenty time sources for the daemon. 
Mine (Ubuntu 22.04 LTS) created seventeen. In my experience, using over a dozen 
server



The way I read the documentation, the number actually used is capped. 
However, I suspect this it he result of converting a server 
configuration for using the pool, to the, newer, pool way, without 
understanding the difference properly.  Using "server" each line would 
only pick one member of the pool specified, and repeated calls on the 
same pool would likely get the same one over and over.



why would there be no guidance for the most prominent distributors of the 
daemon to write sane configurations?


People who package for Linux distributions often aren't power users of 
the packages.  Although the following example isn't NTP, one also gets 
the problem of copying and pasting, which propagates misunderstandings, 
when people who should have read the documentation in detail have just 
copied someone else's solution, with minor tweak.  Where I see this is 
configuration for access internet telephony service providers in the 
Asterisk PABX.  They are invariable half nonsense.


Anyway, I guess the real problem here is that the only maintained 
documentation is reference documentation, and there isn't a good set of 
cook book documentation.



--
This is questions@lists.ntp.org
Subscribe: questions+subscr...@lists.ntp.org
Unsubscribe: questions+unsubscr...@lists.ntp.org

Re: [ntp:questions] ntp pool servers disappear - more data

[David Woolley]

On 26/06/2021 00:12, William Unruh wrote:

Not at all sure what you are suggesting. DNS is a way of translating
names to IP addresses, which your machine MUST use to talk to a remote


As already noted, there is no MUST about it.  I'd put it as low as MAY, 
and it is definitely no more than SHOULD.



machine not on your network. The remote machine has nothing to do with


DNS can be used for local network machines, as well, and this is very 
common.



this. Now some remote machines will as for the name associated with the
IP address of machines sending the remote machine a query, to try to see
if someone is spoofing the IP address, but as far as I know ntpd does
not do that. Takes too much time and would make the time responses
really bad.


 ntpd doesn't care about who is sending it a query, and, in any case 
reverse DNS lookups often provide bad results, which won't match the 
preferred forward lookup, in the real world.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] ntpd on busybox ARM system not keeping time with server

[David Woolley]

On 19/05/2021 01:57, Jakob Bohm wrote:

Perhaps the "tos orphan" option is a better way to make ntpd continue
after loss of all time sources.


This is a pure client configuration.  There is no need for ntpd to 
continue to serve time after the loss of all sources.  The kernel 
software clock will continue to run, using the last available frequency 
correction supplied by ntpd, without the need for any time island 
mitigation measures.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] ntpd on busybox ARM system not keeping time with server

[David Woolley]

On 18/05/2021 12:39, Andreas Schick wrote:

I could safely remove the LCL entries and the server line where it lists own 
IPv4 address of the ARM box


I think it is more accurate to say that you CANNOT safely keep these! 
The self reference is plain wrong.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] ntpd on busybox ARM system not keeping time with server

[David Woolley]

On 18/05/2021 12:26, Andreas Schick wrote:

server 127.127.1.0  # local clock (LCL)
fudge  127.127.1.0 stratum 10   # LCL is unsynchronized


Delete these lines.  As described, this system is not suitable as a time 
server, and including these lines on a pure client can actually 
frustrate synchronisation. This fake server is likely to vote against 
the genuine server.



server  192.168.101.2


This appears to be the machine itself, so it will be voting that's its 
own time is correct.  Delete it.


Windows machines can vary from fair to atrocious as time servers.  A 
workstation running a default configuration of w32time will be at the 
atrocious end.


You should make sure that the ARM starts in the right ball park, by 
either using a file timestamp to record the time at, or close to, 
shutdown, or, as a last resort, setting a fixed time that isn't too far 
from reality.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Incorrect time happens intermittently

[David Woolley]

On 07/04/2021 02:26, Tze An wrote:

[2021-04-06 18:07:25.732] +QNTP: 0,"2021/04/02,10:07:28+32"


Also, what software are you using to process timezones, as "+32" is not 
a real world time zone offset.  It is conceivable that your problem is 
in the software you use to convert UTC to your local timezone, rather 
than in the UTC time provided by NTP.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Incorrect time happens intermittently

[David Woolley]

On 07/04/2021 02:26, Tze An wrote:

I've got incorrect time from NTP server, always 4 days before the actual 
datetime. You can see from the logs below, I got 2nd April on 6th April.

Incorrect time from NTP server
[2021-04-06 18:07:24.841] Clock:Info: Synchronizing time with NTP server...
[2021-04-06 18:07:25.232] NBiot:info:NBiot_Cmd [0x2a]: 
Request[AT+QNTP=1,"0.sg.pool.ntp.org"]
[2021-04-06 18:07:25.732] +QNTP: 0,"2021/04/02,10:07:28+32"

Correct time from NTP server
[2021-04-06 18:09:24.781] Clock:Info: Synchronizing time with NTP server...
[2021-04-06 18:09:25.172] NBiot:info:NBiot_Cmd [0x2a]: 
Request[AT+QNTP=1,"0.sg.pool.ntp.org"]
[2021-04-06 18:09:25.711] +QNTP: 0,"2021/04/06,10:09:27+32"

Has anyone experienced it before?



What software is producing these logs?   In particular, the AT suggests 
this may be using some service provided by a mobile phone.


Please confirm you have at least four,  unrelated servers configured.

Please provide the result of running ntpq against your NTP client, 
assuming that it is sufficiently compliant to support that.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Ntpq.exe memory issue with windows 2019

[David Woolley]

On 28/10/2020 18:01, Uwe Klein wrote:

That does not jibe with "creeping increase of memory taken".

If Win keeps something for reuse ... it should not grow but be reused
later again kept for ...:-)


As I said, I'm not familiar with the way that Windows shows memory 
statistics, but I do know that both on Windows and Linux, people often 
get confused because the OS aims only to keep a small amount of memory 
free.  I was assuming that memory used was total - free, rather than 
reflecting that which could not be released without fatally harming a 
running process.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Ntpq.exe memory issue with windows 2019

[David Woolley]

On 21/10/2020 15:44, Sadique Urf Arbaz Sayyed wrote:

  The process will hold the 24Kb but actually the process is not visible in 
Task manager or anywhere but in RamMap we see that the executed process still 
occupied the Page table memory.


Did you mean 24KB, or is that really 3KB?

Any memory leakage that persists after the process is no longer visible 
in the process table is down to the OS.


As already pointed out page table allocations are unlikely to be what 
that memory is.


I'm not familiar with the details of Window NT family memory use 
reporting, but the only legitimate carry over would be for cached pages. 
 Such pages are optimisations and don't deny memory to other processes.


Both Linux and Windows try and allocate nearly all their memory to 
something.  In the case of the Linux "free" command, it computes an 
available figure which is much larger than the free figure, by including 
memory that can safely be discarded.  If there isn't an OS bug, I wonder 
if you are seeing that effect.


One wild thought: is ntpq or one of its DLLs not position independent 
code?  I could speculate as to why Windows might keep relocated pages 
around in case the code is reloaded at the same address before the page 
gets reused for other reasons.  However, I don't actually know if 
Windows does something like that.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Local Time NTP Server

[David Woolley]

On 15/09/2020 23:20, William Unruh wrote:

the noun is Coordinate


The noun is Time.  The C* is an adjective, coordinated or coordonné.  It 
appears the abbreviation is not in correct word order for either of 
French or English: 
.




___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Local Time NTP Server

[David Woolley]

On 24/08/2020 18:48, William Unruh wrote:

You need it to be
fixed, eg to the stars.


The stars move, and, in any case, most people want solar time, not 
sidereal time.  Even solar time varies throughout the year.  I think the 
"mean" in GMT refers to the fact that midnight is only true on average.


Sidereal time still needs a reference point on the earth that is in 
earth surface coordinates, not galactic ones.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] create charts

[David Woolley]

On 21/08/2020 11:39, thimoo...@gmail.com wrote:

I have a question. how do you make a graph of your ntp server and is that 
possible


What parameter do you want to represent?  Remember that the actual error 
from true time is never known, because, if it could be known, it could 
be made to be zero.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Performance estimation

[David Woolley]

On 16/06/2020 17:11, William Unruh wrote:

The question then is how rapidly the system can respond to an
interrupt,. This at least used to be of the order of a microsecond.
Also, how logd does it take to read the clock with the kernel gettime
routines. They all limit the accuracy of your clock using gps refclock
(and also how long the wire is between the gps unit and the computer)


Also, does the Pi really use the full rate processor clock for real time 
measurement, and is there a slower synchronous logic clock associated 
with the interface  used to receive the external time source?


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Performance estimation

[David Woolley]

On 15/06/2020 15:38, David Taylor wrote:


https://www.eecis.udel.edu/~mills/ntp/html/cluster.html


What is the clock resolution?  If you try and measure jitters that 
aren't several times the resolution, they are not going to be 
particularly valid.


If the hardware clock is almost dead on, and the peak to peak dither is 
just less than the resolution, there will be long periods in which it 
will read as as zero, even though it is actually close to one resolution 
unit.  You could also get cases where dither was very low but read as 
one resolution unit for long periods.  In fact, if it was possible to 
find tune the actual clock oscillator, during an ideal lock you would 
have peak to peak dither, as measured, of one or two resolution units, 
even though the actual phase noise was much less.


(Arguably, a jitter that is less than the clock resolution will result 
in worse time accuracy than one that a few times it, as the clock 
resolution will not be dithered out.


That makes the, normally unrealistic, assumption that the systematic 
error is less than the clock resolution.)


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] time accuracy and stratum

[David Woolley]

On 06/04/2020 11:00, David Woolley wrote:

On 25/03/2020 17:43, Jérôme Perrin wrote:
I need to understand the time accuracy for a client of a given stratum 
(5).

The information for accuracy is retrieved from ntpstat command.


The accuracy only indirectly depends on stratum.

The time is likely to be within +/- root dispersion of the locally 
reported time.  Normally it will be a lot better, but it can also be worse.




That should have been root distance, not root dispersion.

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] time accuracy and stratum

[David Woolley]

On 25/03/2020 17:43, Jérôme Perrin wrote:

I need to understand the time accuracy for a client of a given stratum (5).
The information for accuracy is retrieved from ntpstat command.


The accuracy only indirectly depends on stratum.

The time is likely to be within +/- root dispersion of the locally 
reported time.  Normally it will be a lot better, but it can also be worse.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Trying to reply to an NTP Questions request I get this

[David Woolley]

On 21/08/2019 17:42, David Taylor wrote:
You are not allowed to post to this mailing list From: a domain which 
publishes a DMARC policy of reject or quarantine,


blueyonder doesn't appear to publish any DMARC policy.  Did you actually 
use a blueyonder address as the header From:?  It is also possible that 
they tried one and it broke too many things.


They do seem to have a DKIM signature that includes the subject, so any 
mailing list that rewrites the subject will break the signature.  Most 
mailing lists rewrite the subject, if it doesn't have the list tag in it.


I imagine the list software isn't doing a fully analysis of DMARC and 
DKIM to see if relayed mail will fail, but just looking for specific 
things in the DMARC policy to know whether a failure would be a problem 
for the list.


I'm assuming it is well behaved and rewrites the envelope from so that 
SPF will pass the mail. Otherwise it will be generating SPF failures, 
downstream.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] [META] Trying to reply to an NTP Questions request I get this

[David Woolley]

On 21/08/2019 17:42, David Taylor wrote:

Trying to reply to an NTP Questions request I get this:


You are not allowed to post to this mailing list From: a domain which 
publishes a DMARC policy of reject or quarantine, and your message has

been automatically rejected.  If you think that your messages are
being rejected in error, contact the mailing list owner at %(listowner)s.


As I have no control over my ISP's domain, (blueyonder  co  
uk) could the list owner please resolve this so that I can help the 
person with the issue?  Of course, I get the same message trying to 
contact the list owner by e-mail, hence the post here


I suspect the problem is that, the way they forward mail is such that, 
if they didn't reject a source, a large proportion of the destination 
subsystems would eat, or even worse, bounce the mail, as they will 
interpret the  mail as having been forged.










___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Reference 'sntp' utility: how do you set the destination port number

[David Woolley]

On 21/07/2019 00:26, stua...@longlandclan.id.au wrote:

This implies that someone was naïve enough to assume the destination port 
number was never going to change.


The reference version is almost certainly just that, a reference 
version.  It is there to illustrate how the protocol works, and part of 
that is using the standard port number.  I suspect the maintainers of 
the source code would rather everyone used the full version.


However, although I haven't looked at the source code, I imagine the 
port number is easy to change without any really specific C language 
knowledge, provided you are prepared to recompile it.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Garmin LVC 18x jitter problem

[David Woolley]

On 12/07/2019 11:32, William Unruh wrote:

Do not use nmea as a time source. It it has a long delay ( hundreds of
ms) which depnds on the speed of connection, on length of the nmea
sentance, etc. EAch character takes from 100 of usec to ms to be
delivered, and the lengthof sentences is variable, and the gps sends out
a sentence only when it is not terribly busy ( which depends on how many
sattelites it is trying to handle, etc) It is good for setting the
seconds (sometimes-- if you ask it to send too many sentences it can be
out by more than a second totally messing up your timing).
One sentence (RMA) and a high baud rate make it useful for setting the
second.


I believe the OP is using it for seconds.  The problem is, I think, that 
ntpd will only use it, and only accept the PPS, if it is a true chimer.


Moreover, I think the OP's problem is that the NMEA timing is too 
repeatable, so its root dispersion is too small to cover the static 
error, although, if they had a lot of jitter, they would then advertise 
a root dispersion that doesn't give credit to the PPS accuracy.


This is a degenerate case of root dispersion, as it is measured at the root.

Rather than setting as separate error bound for the NMEA, is there any 
option to set a static offset separately for the NMEA.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Time server question

[David Woolley]

On 21/06/2019 12:26, Thomas Laus wrote:

Will either isolation solution have direct access to the computer
CPU?  The GPS clock will need the ability to directly adjust the
frequency of the CPU to achieve expected results for a Stratum 1
serve


I'm not aware of anything in ntpd that directly adjusts the CPU 
frequency and there generally isn't any fine grained way of doing that. 
ntpd normally works by adjusting how many cycles of a fixed frequency 
represent a certain time period, and that is a software operation.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

[ntp:questions] [META] Please Report Dissolved Boxes Spam to Google.

[David Woolley]

Recently I've been religiously reporting the Case Solutions spam that is 
appearing on the newsgroup side of this combined list and newsgroup, 
with no effect.


The spam is being injected via Google, but their groups-abuse email 
address appears to be broken, so there is only the web interface 
available to submit reports, and it does not allow free text reports.


My guess is that Google will only care when multiple independent reports 
are seen for individual postings, which will not typically happen 
because this is a niche newsgroup.


Could I therefore implore anyone receiving this, who has a Google 
account, not to kill file Case Solutions, but to also report all their 
spam to Google.


If an anyone knows how to get a real human, with at least the half brain 
needed to recognize this as serious abuse, at Google onto solving the 
case, even better.


I paraphrased the spammers tag to get past existing kill files.

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Hoes a IoT device discover the NTP era with NTPv4?

[David Woolley]

On 11/03/2019 19:59, Nelson Bolyard wrote:


But 2036 is fast approaching.  I want to know how, using NTPv4, a device like 
mine can figure out that it is no longer in era zero.  Once it figures out the 
era that it is in, the rest should be easy.


Assume that no date is before the design date of the product, and 
interpret anything lower than that as era 1.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What is the mean of below output?

[David Woolley]

On 06/08/18 08:16, ashishchugh@gmail.com wrote:

3- Is there is any way through i can determine that what is the currently 
difference between my local system and ntp server.


No.  If you could, NTP could modify the local time to make the value zero.

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] How can i make sure that how much time ntp is adjusting one day

[David Woolley]

On 05/08/18 07:15, aashish.ch...@fonantrix.com wrote:


i enable the ntp, how can end of the day i can get the total time which ntp 
adjusted.


NTP adjusts frequency, not time.  It doesn't predict what the internal 
clock would show if it hadn't been adjusted


when i execute ntpstat command i can see following output

synchronised to NTP server (169.254.169.123) at stratum 4
time correct to within 9 ms
polling server every 8 s


I'm not sure what is being measured to produce the 9ms figure.  It seems 
suspiciously
low for a good estimate of the error bounds for a such a high stratum 
client.


If you are polling the server every 8 seconds you have overridden the 
default settings in a way that is not recommended, is likely to cause 
sub-optimal time keeping, and is likely to get you blacklisted by many 
servers.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Difference between offset (ntpq -p) and offset (ntpq -crv)

[David Woolley]

On 01/08/18 19:32, David Taylor wrote:
Would I be right in thinking of the "*" line as simply being the offset 
from that particular server, and the "system" variable as being the 
offset from some virtual internal clock which ntp has as its best 
estimate of the correct time (e.g. UTC).


That cannot be correct, as the other side of the comparison is that best 
estimate of system time and the result would always be zero.


Without digging into the code, my guess is that it is the offset from 
the weighted average of the last reading from all the true chimers.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] This is project have any SLA and does it meets th PCI and DSS compliance

[David Woolley]

On 12/05/18 18:55, bhuvaneshwara...@powerupcloud.com wrote:

We are going to use these pool servers for our production env, so we need to 
know that you have any SLA for uptime and does this project meets the PCI and 
DSS compliance?


There is no service level agreement.  If you want a service level 
agreement, you will need to use a commercial source, although, 
ultimately commercial sources are likely to depend on satellite 
navigation systems that are unlikely to have service level agreements 
with anyone, except, possibly, their own military.


Do you mean PCI DSS, rather than both PCI and DSS?  I.e. Payment Card 
Industry Data Security Standard?  Nowhere is payment card data requested 
or stored. There is no-one to pay for any certification that that is the 
case.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Who Owns or Administers this Group?

[David Woolley]

On 05/02/18 20:09, Steve Sullivan wrote:

[No question.  Assuming the subject should have been in the body.]


Steve Sullivan | Client Support & Development
Network Time Foundation



This "group" is actually two "groups" gatewayed together.  I am reading 
this on the USENET side of the gateway, and, as is usual for USENET, 
no-one owns or administers that.


Details of the mailing list side can be found at: 



___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] false ticker after GPS coldreset

[David Woolley]

On 01/11/17 11:46, valizadeh...@gmail.com wrote:

local lcok is there because my system is not connected to internet and i need 
to have the hwclock to keep the time during power-offs,i have disabled the pre 
installed fakeclock and used an I2c connected battery backed up RTC chip.


Local clock has nothing to do with the RTC chip.

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What I've learned about sun clocks and ntp

[David Woolley]

On 03/08/17 10:56, rahl...@gmail.com wrote:

If I understoodhttps://tools.ietf.org/html/rfc5905  correctly,
NTP stratum-1 servers have to provide UTC time and cannot use
an alternative timezone.


If you can live with mean solar time, rather than sundial time, all 
modern Unix-like systems can do this once calibrated.  Although you 
might think in terms of of quarter hour timezones (Nepal is on a quarter 
hour), the Olson already has mean solar time defined for many places, 
for dates earlier than the introduction of global time standards 
(although there may be some limitations on using that information with 
times before the Posix epoch).


You can create your own solar timezones by compiling suitable tables for 
your own location.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What I've learned about sun clocks and ntp

[David Woolley]

On 03/08/17 13:02, William Unruh wrote:

UTC is mean solar time at Grenwich in London


UTC was never this.  GMT used to be, but I think is now UT1.  UTC has 
only ever ticked at the TAI rate, with leap seconds.


s/Gre//

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] NTPQ -P shows both IP and DNS name (parsing problem)

[David Woolley]

On 20/06/17 14:55, roman.mescherya...@gmail.com wrote:

-193.11.114.43 (tor1.mdfnet.se)

See the line starting with “-193.11.114.43 (tor1.mdfnet.se)”

This strange peer breaks extracting fields by index. For the above example it 
extracts “(“ as “refid” value instead of “75.17.28.47” and “29.118” as “offset” 
value instead of “-0.185”.



I think you are expected to use the relevant management request 
directly, rather than parse output intended for humans.  That would 
avoid process startup, filtering, and DNS costs.




Is this behaviour a bug or a feature?

"
Whilst I haven't looked at the code, I wonder if tor is Totally Off the 
Record", in which case it is quite likely it doesn't reverse resolve 
correctly.  My guess is that it is displaying the information in this 
form because the reverse resolved name doesn't match the one used, and 
therefore indicates a possible security issue.


In this case, it looks like it reverse resolves to a non-existent domain 
name.




___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] NTP under AIX?

[David Woolley]

On 16/05/17 16:37, Greg Moeller wrote:

Has anyone come across the advisability of running an enterprise-wide NTP 
server under an AIX LPAR?
We're currently running NTP on old Intel hardware and the company policy is to 
refresh hardware on a regular basis.
It seems a waste to buy several new servers if we could just put the NTP 
service on an AIX LPAR.



If you are talking virtual machines, ntpd should always be run on the 
host.  Any use on a guest should only be as a leaf node.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] what is refid in ntpq -pn

[David Woolley]

On 06/05/17 22:24, Hans Mayer wrote:



I always thought "refid"  for command "ntpq -pn" shows the next upstream server for the 
remote server listed below "remote". But now I have my concerns.



Officially, it is an opaque 32 bit identifier for the selected upstream 
peer of the peer.  You are not supposed to assume it has any more 
meaning, even though, on pure IPv4 systems, it is derived in the way you 
describe.


Note that the selected peer is not the only source of time; it is just 
the one used for the quality parameters.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] NTP server - Number of received petitions.

[David Woolley]

On 08/03/17 10:53, Micron wrote:

Recently, I've added my NTP server in the pool and I'm looking for the way to 
know the number of petitions received from clients.


Please could you indicate where petition is defined in the context of 
NTP.  I have never seen it used before.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] How common is LI=3 - solved.

[David Woolley]

On 05/02/17 00:22, Robert Scott wrote:

 I hope to achieve a
frequency accuracy of 5 PPM.  Once that measurement is made, I store
it for subsequent use in my app.


The equipment doesn't have a very long service life and is in a 
temperature controlled environment, as I think both ageing and thermal 
effects can violate the 5ppm limit.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] How common is LI=3 - solved.

[David Woolley]

On 03/02/17 23:10, Robert Scott wrote:

 But with other servers, the first
response came back good, the second response came back LI=3, the third
and fourth responses never came back at all.  (recvfrom() has to be
aborted).

Is this possibly a defense mechanism against a DOS attack?



See , in 
particular the section on Kiss-of-Death.  Your code should be 
recognizing these and raising an alarm.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Writing the drift file

[David Woolley]

On 08/03/15 11:01, Mike Cook wrote:


Current practice appears to be to open and write each new value to a temporary 
file then unlink the old and rename the temporary file.
This means that a whole new file structure including new blocks is created each 
time and the old one freed up. So even in the worst case where the same blocks 
are allocated on alternated file updates, the effective life of the device is 
doubled.


I believe that all but the crudest flash devices use copy on write at 
the physical level, as part of their wear balancing strategy.  Updating 
directories as well a the file will increase the wear!


This is complicated because the unit of writes is typically much larger 
than allocation unit, so one may sometimes find that, with write 
caching, there are less physical writes than implied by the complexity 
of the update.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] chrony as a server

[David Woolley]

On 23/02/15 21:23, William Unruh wrote:

manual corrections are probably good to 1 sec. to get 1 sec at 2ppm is
about 5 days per measurement or 10 days altogether.



It's a long time since I did this, but 200ms is more like it (might have 
been 100ms).  You need  digital clock that is, itself, synchronised, and 
you match the rhythm of the ticks, then let one of them actually move 
the return key.


You have to really believe in setting accurate time not just be 
following some instructions, to do this.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Pool server gone wild

[David Woolley]

On 20/02/15 08:45, Roger wrote:

 http://www.pool.ntp.org/scores/90.155.73.34

How does one alert an operator that their server is sick?
Checking back through my peerstats I see that last entry
which was okay was 2015-02-16 15:08:56.



You could run whois on the address and contact the contact address 
given.  It is a small ISP (AAISP), so there is a good chance that the 
message will get through.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] chrony as a server

[David Woolley]

On 15/02/15 22:40, Rob wrote:

it is tracking very nicely


Tracking what?

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] NTP offset doesn't change.

[David Woolley]

On 10/02/15 05:15, catherine.wei1...@gmail.com wrote:

Hi, I'm using the ntpd to sync time. When I change the current date for exampe to 
0210020215 (2015-02-10 02:02), the actually current time is 2015-02-10 03:02, then I run 
ntpq -p for several times, the offset doesn't change at all.


Garbage In Garbage Out.

ntpd is intended in environments where time behaves like time.  It 
should actually abort in the case you describe, as something has clearly 
broken so badly that it is no longer safer to try and discipline the clock.


People often try this as a test, but it is not a valid test.

Basically, up to a lower limit (default 128ms, but significantly 
increased by -x) ntpd will slew to correct.  Above that limit, up to a 
second limit (which I seem to remember to be 10 minutes) it will wait 
about 15 minutes to confirm that it really is seeing a step change, then 
restart the initial synchronisation and step.  Above the second limits, 
it should abort, except if you enable an option to allow one such step 
and it is the first one.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Shared PPS source/Multiple PPS sources

[David Woolley]

On 07/02/15 08:24, Rob wrote:

And a side question: Is it the GPS module that calculates when the PPS goes 
active? Is this signal compensated for the time it takes the signal from the sats 
in the module, or on the SV?

Yes, the module calculates the position and time fix from the signals of
at least 4 SVs, and then outputs the PPS related to the calculated time.


Which is the key to how GPS does the P.  GPS receivers need to know 
where the satellites were when the signal was launched, and the orbits 
are specified in terms of the satellite times.  It then has to work out, 
from the relative delay times, from at least four satellites, the 
current time and three spatial coordinates (four unknowns).  I'm not 
sure it actually needs the time part of the solution (it probably does 
for detailed corrections) but that is going to drop out of solving for 
the spatial position.


What the satellites do do is to to transmit some additional information 
about ionospheric propagation, which allows for receiver to compensate 
for the signals travelling slower than the speed of light in vacuo.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Mitigating the ::1 spoof vulnerability

[David Woolley]

On 06/02/15 12:17, Marco Marongiu wrote:


I'm referring to this one in particular: ::1 can be spoofed on some
OSes, so ACLs based on IPv6 ::1 addresses can be bypassed.

Debian Squeeze doesn't have a patched package available in the
squeeze-lts series yet. On those clients would a restriction like

restrict ::1 ignore

mitigate the vulnerability?



Sounds more like you need to fix the firewall.  Firewalls should not 
allow incoming source address 127.0.0.1 and internet level firewalls 
should not allow private use only source addresses.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Leap second to be introduced in June

[David Woolley]

On 26/01/15 17:11, William Unruh wrote:

physical principle ( the frequency of oscillation of a cesium atom in a

 XX

certain transition) and the rotation of the earth. It used to be defined

  ^not




It's a quantum spin flip.  There is no physical oscillation. Not that 
that really changes anything about UTC or TAI.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Timekeeping on Windows 2008r2 VM on Linux QEMU/KVM

[David Woolley]

On 22/01/15 00:02, William Unruh wrote:

I believe ntpdate simply sets the clock. No slewing. It could not, since
slewing must assume that the program is running for a while so it
can switch off the slewing. ntpdate just runs once.


The adjtime takes an offset and then slews at +/- 500 ppm until that 
offset has been cleared.  It doesn't need the requesting process to 
still be running.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Leap second to be introduced in June

[David Woolley]

On 21/01/15 10:39, Mike Cook wrote:

   I couldn’t find a definition of a monotonous function.


It's an obvious mis-choice of words by someone whose name suggests they 
aren't native English speaker.  It clearly is intended to mean 
monotonic.   See https://en.wikipedia.org/wiki/Monotonic_function for 
a definition.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Leap second to be introduced in June

[David Woolley]

On 19/01/15 12:15, Mike S wrote:


You clearly misunderstood TF.460, because you still have it wrong. There
is no discontinuity, the two scales merely count time differently. This
is how the time of the next leap second will be enumerated in each:



You are relying on an appendix that deals with representation of dates. 
The main part of the standard is worded in terms of their being missing 
seconds.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] I don't understand Restrict statements!

[David Woolley]

On 19/01/15 09:55, David Taylor wrote:

restrict source notrap nomodify nopeer Xquery
restrict 192.168.0.0 mask 255.255.255.0 Xpeer


If X is not no the permission is enabled, and you do not need to 
specify it.  restrict lists restriction.  Everything unrestricted is 
allowed.  Only one restrict line matches.  Broader matches are not 
refined by narrower ones.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Leap second to be introduced in June

[David Woolley]

On 15/01/15 07:56, Terje Mathisen wrote:


Did we have a leap second last June? Or did you intend to check for 2015?


Oops.   I did get it right in the dry run, but not in the run I actually 
used:


david@dhcppc4:~$ TZ=/usr/share/zoneinfo/right/UTC date -d '30 June 2015
86400 seconds'
Wed Jul  1 00:00:00 UTC 2015

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Leap second to be introduced in June

[David Woolley]

On 14/01/15 16:37, Terje Mathisen wrote:

The calls I'm thinking of are those you make to convert an OS-supplied
time_t (file) system timestamp to YMDHMS etc.


Those calls have no need to be in the kernel, and they are not in 
Unix/Linux systems.


I.e. even Windows (which uses a seconds-based timestamp with 100 ns
resolution) has calls exactly like that. Those are the calls that would
need to be updated in order to work in (TAI * 1e7).


There is no need to update them on Debian Wheezy, and probably most 
Linux and modern Unix systems.



david@dhcppc4:~$ TZ=/usr/share/zoneinfo/right/UTC date -d '30 June 2012 
86399 seconds'

Sat Jun 30 23:59:59 UTC 2012
david@dhcppc4:~$ TZ=/usr/share/zoneinfo/right/UTC date -d '30 June 2012 
86400 seconds'

Sat Jun 30 23:59:60 UTC 2012
david@dhcppc4:~$ TZ=/usr/share/zoneinfo/right/UTC date -d '30 June 2012 
86401 seconds'

Sun Jul  1 00:00:00 UTC 2012

I haven't double checked that this is not a trick in the user interface 
code, and they don't work for future time, until the leap second has 
been declared and distributed (in the tzdata files):


david@dhcppc4:~$ TZ=/usr/share/zoneinfo/right/UTC date -d '30 June 2014 
86400 seconds'

Tue Jul  1 00:00:00 UTC 2014

The inability to record future civil times in an efficient format is one 
of the disadvantages of TAI.




___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Leap second to be introduced in June

[David Woolley]

On 12/01/15 16:10, Marco Marongiu wrote:

If so, does it also mean that it would do the same when you disable the
kernel discipline by adding a disable kernel in ntp.conf?


(Or by trying to disable stepping.  A lot of people seem to run systems 
that are incompatible with the use of the kernel discipline.)


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] UK pool server denying access

[David Woolley]

On 30/12/14 08:21, Mike Cook wrote:

This guy needs bouncing


The abuse address for the payload web site's hosting company is: 
mailto:ab...@eukhost.com


That for the company that hosts the submission account is: 
mailto:groups-ab...@google.com


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What to do for clients less than 4.2.8?

[David Woolley]

On 22/12/14 04:02, Paul wrote:

And yet people apply critical monthly patches from Microsoft and Oracle all
the time without running them through dev and q/a.


Not on business critical servers.  They may well apply them to general 
purpose desk top machines, but even then, if they don't have enough 
diversity, that can be a serious risk.


Also, what happens here is more akin to service pack, which is even more 
likely to get extensive lab testing.


I'm not sure if I've had a Microsoft update break anything in my 
non-critical system use, but I've certainly have had false positives 
from virus checker updates causing damage which wasted a hour or two on 
a  my home system, but if it had affected an important component on a 
critical server, or even on all the company workstations, it would be 
disastrous for the company.


Many businesses operate local repositories of Microsoft updates, not 
just to reduce bandwidth.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What to do for clients less than 4.2.8?

[David Woolley]

On 20/12/14 22:01, Rob wrote:

David Woolley david@ex.djwhome.demon.invalid wrote:

On 20/12/14 19:58, William Unruh wrote:

Is it an ntp packet (ie a time exchange packet)? is it a control packet
(eg ntpq type packet?) or what?
Ie, unless you use crypto, these two look like they might be dangerous.


Both routines only process NTP type 6 packets, i.e. nptq.


But is that before or after those packets are filtered by restrict?

ctl_putdata is sending the response (my guess is the attack is monlist 
based), so it is definitely after the filter.  configure is a fairly 
complex command processing option, so, although I didn't check the code 
in detail, I would be most surprised if it wasn't after the filter.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What to do for clients less than 4.2.8?

[David Woolley]

On 20/12/14 20:54, A C wrote:

Ok, so the remaining uncertainty is whether some of the crafted packets
can be the response packets for a normal time exchange or if they're
only query/config packets.  The advisory isn't completely clear on what
types of packets can cause the buffer overflows.


ctl_putdata handles the responses to ntpq type control packets. 
configure is the action routine for a particular control type request. 
They are both in ntp_control.c, whose first four lines are:


/*
 * ntp_control.c - respond to mode 6 control messages and send async
 * traps.  Provides service to ntpq and others.
 */

I didn't check the encryption one, as casual users don't use encryption. 
 It may well turn out to be the encryption used for control packets.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What to do for clients less than 4.2.8?

[David Woolley]

On 21/12/14 10:48, Rob wrote:

People say disable crypto but there is no clear direction in the docs
on how to do that.  There is no crypto off or disable crypto config
directive at first glance.  So how is this done?


I would assume by not enabling it.

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Restrict statements and the pool directive

[David Woolley]

On 21/12/14 11:24, Rob wrote:

Anyway, I consider it a bug.  I don't want to lift restrictions to
arbitrary systems selected from a pool.  So, out went the pool command.


Why do you want to specify pool servers if you want to restrict their 
use so that you cannot use them?


When people say lift restrictions here, they mean lift those which 
prevent the use as a server, not lift all restrictions.  Unless you hare 
using a blunderbuss approach to restrictions, or using ignore you 
should not be blocking the access needed to act as a client.


I think the original question on this thread is much more likely to be 
due to a DNS problem than one in using restrict, as I cannot see 
anything in those restrictions which would prevent client access to the 
pool.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What to do for clients less than 4.2.8?

[David Woolley]

On 21/12/14 11:38, Rob wrote:

David Woolley david@ex.djwhome.demon.invalid wrote:

On 21/12/14 10:48, Rob wrote:

People say disable crypto but there is no clear direction in the docs
on how to do that.  There is no crypto off or disable crypto config
directive at first glance.  So how is this done?


I would assume by not enabling it.


Ok, but in that case why the worry about the millions of vulnerable
servers on the internet, I think most users who just want to get and


Paranoia?  Security alerts are generally not that explicit (and this one 
is actually unusually explicit) because they provide information to the 
hackers.



serve time don't spend the week of time needed to get the crypto working
and to coordinate with other servers doing the same.

So for now I presume it is on by default...  also because of what I saw
in the OpenSUSE example config.  (or would the keys config directive
be the magic enable crypto directive?)


There are only two places where crypto_recv is called.  One is 
definitely only active if autokey has been explicitly configured.  The 
other is only active for broadcast clients and the comments imply that 
it is only used for autokey, but it does seem possible that it is the 
remote side that decides this (I didn't follow the code any deeper); it 
is in the initial broadcast client handshake.


I'm using 4.2.7p333, rather than the latest 4.2.7 source code.

Carefully crafted in alerts generally means that the data has to look 
like the address of some instructions and those instructions, with the 
exact memory layout under which that instance is running.  It also 
normally assumes that the machine doesn't have stack execution 
permission disabled for ntpd.




___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Restrict statements and the pool directive

[David Woolley]

On 21/12/14 20:10, Rob wrote:

What I got from the documentation is that without nopeer a server
could setup a peer association.  I don't like that.


No. Without nopeer, a *client* can't set up a peer session.  If you are 
using a system as a server, it cannot cause you more disruption than if 
it peered itself with you.


The problem here is that the exact significance of being a peer isn't 
well documented.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What to do for clients less than 4.2.8?

[David Woolley]

On 20/12/14 09:22, Martin Burnicki wrote:



As far as I understand the reports on bugzilla the main vulnerabilities
are in functions where signed packets (symmetric key or autokey) are
received/checked, or dynamic/remote configuration via ntpq and/or ntpdc
is enabled, which, as far as I know also requires some sort of crypto
top be enabled.



One might be in a pure status enquiry, so you may have to set noquery.

In any case, except possibly for people using encryption, and maybe not 
even them, these affect neither client nor server mode, only remote 
management.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] What to do for clients less than 4.2.8?

[David Woolley]

On 20/12/14 19:58, William Unruh wrote:

Is it an ntp packet (ie a time exchange packet)? is it a control packet
(eg ntpq type packet?) or what?
Ie, unless you use crypto, these two look like they might be dangerous.


Both routines only process NTP type 6 packets, i.e. nptq.

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] ATOM driver not working on Linux

[David Woolley]

On 14/12/14 09:33, David Taylor wrote:

On checking the config.log I see that timepps.h is missing:


Please read before posting.  This has been extensively discussed over 
the last week.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] NTP PPS, part 2 ;)

[David Woolley]

On 12/12/14 12:40, Martin Burnicki wrote:

Harlan Stenn wrote:

Martin Burnicki writes:

IMO the best approach would be to detect this at runtime.


That means we'd need a header file...


It shouldn't be a problem to add this to the NTP code base.


NTP doesn't control this interface.  The de facto interface is defined 
by the kernel code.



___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] NTP PPS, part 2 ;)

[David Woolley]

On 12/12/14 16:28, Paul wrote:

On Fri, Dec 12, 2014 at 9:47 AM, David Woolley 
david@ex.djwhome.demon.invalid wrote:


NTP doesn't control this interface.  The de facto interface is defined by
the kernel code.



I don't understand this.

My misunderstanding.  I thought this was doing the job of linux/pps.h. 
As it is not.  I would agree that it is not the job of the OS.  It 
should either be in a separate package, which is buildrequired by NTP 
packates, or it should be snapshotted into ntpd itself.


It's just another application library.

___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Re: [ntp:questions] Red Hat vote for chrony

[David Woolley]

On 09/12/14 07:06, Charles Swiger wrote:

Yes, I also find it a bit surprising than modern desktop CPUs and GPUs
are willing to run right up to their thermal trip points of ~80 C or so
rather than bump up fan speed a little more to keep them more around 50 C.


Cost engineering, for what is a throwaway product with a fashion life of 
less than 3 years.  Also,


Older systems tended to use more aggressive cooling, especially laptops.

Well, smarter firmware and Hall effect sensors to measure fan speed means
you can spin the fans more slowly than if you needed to apply 40% minimum
speed just to be sure that the fan would spin up from idle.


Variable speed fans are available now and motherboards support them. 
Tachometer outputs have been around for a long time, and because fans 
use brushless DC motors (i.e. synchronous motors with electronics to 
generate the AC) they all have to have rotational phase detectors 
(although it may be done by monitoring the coil current, rather than a 
separate sensor).  Heavy use of these tends to be associated with quiet 
PCs, so the BIOS may well be set for minimal cooling, rather than 
keeping a low temperature.


___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions