NTPD not keeping time

[Chuck Swiger]

On Nov 30, 2006, at 12:16 PM, Kris Anderson wrote:
> Here's the output from ntpq.
>
> webdev# ntpq -p
>      remote           refid      st t when poll reach
>  delay   offset  jitter
> ====================================================================== 
> ========
>  time-a.nist.gov .ACTS.           1 u  485 1024  377
> 78.454  4307608 923174.
>  india.colorado. .ACTS.           1 u  491 1024  377
> 22.918  4307064 922326.
>  lerc-dns.grc.na .INIT.          16 u    - 1024    0
>  0.000    0.000 4000.00

Your clock is off by a little over an hour; while ntpd can correct  
very large offsets, doing so takes a long time.  Kill ntpd, re-run  
"ntpdate -b", double-check that your clock is sane, and then re-start  
ntpd.

You should also note that your third ntp server is not answering  
queries, so you should try finding some other ntp server to use.  Are  
you providing time syncronization from this machine to other hosts,  
or are you only running as a standalone client?

If you are not providing time sync to a large subnet, please consider  
using stratum-2 servers or the NTP pool, ie, pool.ntp.org, or more  
specific regional parts, such as 0.us.pool.ntp.org,  
1.us.pool.ntp.org-- this is assuming from your IP that you are  
located in the US, otherwise choose the appropriate country code for  
where-ever you are.

-- 
-Chuck