Re: GNU/Linux System Clock Drift

On 10 May 2025 16:09:58 GMT, Stéphane CARPENTIER wrote:

 
The hardware drifts, yes, there is no way around. But the system can
estimate its drift and compensate for it. So, once again, unlike your
beliefs, my hardware drift but my system doesn't.
 

Both the hardware and system clocks will drift.

The hardware clock is maintained by the BIOS.

The system clock is maintained by counting the timer interrupts
and there are adjustments that can make this more accurate but
it will still drift.

That's why there is synchronization, usually, but not only,
with NTP time servers.

The issue becomes how often does the system clock need to be
synchronized.  I do this at each boot using openrdate.  The
distros probably do it much more often.  I could also do it
more often but I choose not to do so.

There is no need for me to sync frequently.  File time stamps
will never be more than a second or two off and that is of no
consequence.

But the important fact is that _I_ choose to do it.  I do not
follow any distro but make my own decisions.

 
Of course my system does include leap seconds.
>

What does this command output:

TZ='/usr/share/zoneinfo/GMT' date +%s --date="May 10 2025"

It should output "1746835227."  If not then your system does
not include leap seconds.





--
Systemd: solving all the problems that you never knew you had.