I am running ntpd using ntp-4.1.2-0.rc1.2 under RedHat 9 running dual 2.8GHz
Intel Xeon processors in a Dell Power Edge 1600SC server,

/etc/ntp/drift stabilizes to around 100.  I understand this to mean 100 part
s
per million = 9 seconds/day = 53 minutes/year.  That strikes me a large, eve
n
for a cheap digital watch.

Is this some kind of an artifact of the kernel or of ntp?  If not, is this a
n
acceptable drift for the hardware clock?  Is it something I should worry abo
ut?
Does it bode trouble for me as the board ages?  Should I complain to Dell?

[ Post a follow-up to this message ]

anonymous <neAwsAAA@daCd.orAg> wrote in message news:<slrncc77pp.ga6.neAwsAAA@tad.dad.org>..
.
> I am running ntpd using ntp-4.1.2-0.rc1.2 under RedHat 9 running dual 2.8G
Hz
> Intel Xeon processors in a Dell Power Edge 1600SC server,
>
> /etc/ntp/drift stabilizes to around 100.  I understand this to mean 100 pa
rts
> per million = 9 seconds/day = 53 minutes/year.  That strikes me a large, e
ven
> for a cheap digital watch.
>
> Is this some kind of an artifact of the kernel or of ntp?  If not, is this
 an
> acceptable drift for the hardware clock?  Is it something I should worry a
bout?
> Does it bode trouble for me as the board ages?  Should I complain to Dell?

Someone just posted a response to a similar issue with Dell's running
SMP on Linux. It involved replacing the BIOS. See the message for
details.

Danny

[ Post a follow-up to this message ]