Re: [chrony-users] Fwd: Chrony malfunctioning at beaglebones

[ Thread Index | Date Index | More Archives ]

On Tue, 3 Nov 2015, Miroslav Lichvar wrote:

On Tue, Oct 27, 2015 at 01:40:10PM +0000, Nuno Gonçalves wrote:
On Thu, Oct 8, 2015 at 7:58 AM, Miroslav Lichvar <mlichvar@xxxxxxxxxx> wrote:
If you can, please test the new code in git and let me know if it
doesn't work as expected.

All works fine but...

Oct 27 13:37:31 white chronyd[635]: RTC time before last driftfile
modification (ignored)

This message appears if I restart Chronyd on a running system. Well,
RTC time is before last driftfile, but we are not restoring from
neither of them, because the system clock is ahead of both.

Not sure If this will create any user confusion.

I think it depends on what is the user expecting from the -s option.
If someone decided to fix system clock that's ahead of true time by
setting the RTC and restarting chronyd with the -s option, this
message would explain why it didn't work, even though it worked with
older versions.

Normally, the -s option should be used only on first start. A SysV
init script could be modified to do that, but systemd for instance
doesn't seem to allow different commands for first start and other
starts or restarts.

Of course you could have systemd call a script which determined whether this
was a first start or not and made the -s call or not. (which is essentially what the sysV startup script would do)

Mail converted by MHonArc 2.6.19+