Production System loses an unusually large amount of time..... VERY strange

I have a system installed at a client site, using an Intel D410PT Mainboard (Mini ITX) with a built on Atom 1.8gHz processor (Only 5 extensions here…) and it seems to loose 2 to 3 seconds PER MINUTE on the system clock, is there something I’m missing? Just wondering if anybody else has experienced this issue.

Here’s a snippet of a log I have going for the ntp updates – Notice how much adjustment is needed PER MINUTE… I know running this once a minute is a “Band Aid” fix, but it’s all I can come up with for now until I can find what the root of the issue is…

2 Aug 13:03:01 ntpdate[16949]: ntpdate [email protected] Sat Dec 19 00:58:17 UTC 2009 (1)
2 Aug 13:03:09 ntpdate[16949]: step time server 17.151.16.20 offset 2.728490 sec
2 Aug 13:04:01 ntpdate[17035]: ntpdate [email protected] Sat Dec 19 00:58:17 UTC 2009 (1)
2 Aug 13:04:05 ntpdate[17035]: step time server 17.151.16.23 offset 2.648494 sec
2 Aug 13:05:01 ntpdate[17123]: ntpdate [email protected] Sat Dec 19 00:58:17 UTC 2009 (1)
2 Aug 13:05:04 ntpdate[17123]: step time server 67.18.187.111 offset 2.732872 sec
2 Aug 13:06:01 ntpdate[17251]: ntpdate [email protected] Sat Dec 19 00:58:17 UTC 2009 (1)
2 Aug 13:06:05 ntpdate[17251]: step time server 67.18.208.240 offset 2.773852 sec

-Andy

Hello,

I have seen this in two cases: The CPU is completly overloaded or the CMOS Battery was empty. In very rare cases it was actually a bad mainbord…