[ntp:questions] Re: NTP for dummies

Richard B. Gilbert rgilbert88 at comcast.net
Sun Oct 2 13:34:54 UTC 2005

Ephraim F. Moya wrote:

>On Sat, 01 Oct 2005 23:09:47 +0100, david at djwhome.demon.co.uk (David
>Woolley) wrote:
>>In article <433E8CD6.7000904 at gis.net>, mayer at gis.net (Danny Mayer) wrote:
>>>ntpd on Windows does set the CMOS clock, at least it's supposed to.
>>Not in version 4.2.0 and I would consider it a bug for it to set
>>in a later version as it would be redundant, would risk a conflict [A]
>>with the kernel code that already does this, and might fail to
>>implement the policy that only the low order part of the hardware
>>clock is adjusted, to allow people to dual boot Windows, which
>>requires the hardware clock to be in wall clock time.
>>This code in the linux kernel (2.4.26) already ensures that the CMOS clock
>>tracks a synchronised software clock (arch/i386/kernel/time.c):
>>       /*
>>        * If we have an externally synchronized Linux clock, then update
>>        * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be
>>        * called as close as possible to 500 ms before the new second starts.
>>        */
>>       if ((time_status & STA_UNSYNC) == 0 &&
>>           xtime.tv_sec > last_rtc_update + 660 &&
>>           xtime.tv_usec >= 500000 - ((unsigned) tick) / 2 &&
>>           xtime.tv_usec <= 500000 + ((unsigned) tick) / 2) {
>>               if (set_rtc_mmss(xtime.tv_sec) == 0)
>>                       last_rtc_update = xtime.tv_sec;
>>               else
>>                       last_rtc_update = xtime.tv_sec - 600; /* do it again in
>>60 s */
>>       }
>>[A] not all versions of linux support /dev/rtc, which would be necessary, but
>>possibly not sufficient, for user space code to update the clock atomically.
>You seem to be describing linux only. ?? Is the same true for WinXP?
>If not, what is true for WinXP?
>ntp seemed a good solution since I thought I could forget about
>calibrating it once I install it.
>It seems that ntp may not be what I want. Can somebody recommend a
>program for WinXP that will consult a time server once an hour (or so)
>and set my hardware clock. I'd like it to be as independant as
>possible since I envision it running in the background with no
>attention from me.
>FWIW the hardware clock on my computer drifts about 1.5 seconds per
>day. I can set it by hand but I'm not fast enough to get within a
>second and I don't want to mess with it normally.
I think you may be confused about the clocks on Windows.   Yes, there 
are two.   Windows maintains a software clock; it's what displays the 
time at the right hand side of the tool bar.   There is a also a 
hardware clock (CMOS) that is used only to set the software clock when 
the machine is rebooted.  Neither one wins any prizes for accuracy.   
The CMOS clock is updated with the current time from the software clock 
when the system is shut down.  NTP disciplines the software clock.

Instead ntpd, you might use the sntp client available in the reference 
implementation or you might use W32time, Microsoft's 
non-standards-compliant version of sntp.  Any of these three options 
should keep the software clock within fifty milliseconds or better.   
Ntpd should do the best job but it may be overkill with respect to your 
I use W32time on all my Windows 2000 and XP systems; it's the easiest 
way.  Bring up the command prompt and type:
net time /setsntp:ntpserver
net time start
and you're done.   Replace "ntpserver" with the IP address of the system 
you wish to use as a server.  W32time queries the server several times a 
day and tweaks the software clock to bring it into line.

Google for "The Windows Time Service" by Shala Brandolini and Darin 
Green.  It's a Word document that describes the service in some detail.

More information about the questions mailing list