[ntp:questions] Re: Error returned from NTPTRACE 4.2

Ronan Flood ronan at noc.ulcc.ac.uk
Thu Apr 27 16:51:55 UTC 2006


"Conan" <conan at nortel.com> wrote:

> We have 3 internal GPS time sources; CA, US & Europe. The North American 
> units are both Spectracom , EU is very old TrueTime. I'm seeing error 
> "***Request timed out" when our latest Solaris & Linux desktops try to run 
> NTPTRACE ver4.2 against either Spectrocom.
> 
> # ntptrace
> localhost: stratum 3, offset -0.000148, root distance 0.002669
> zcars0vs: stratum 2, offset 0.000222, root distance 0.002522
> 47.129.242.109: timed out, nothing received
> ***Request timed out
> 
> Seems odd that NTPTRACE is returning the IP, since all ntp.conf references 
> use DNS aliases.

ntptrace doesn't use ntp.conf, it reads the IP addresses from the
NTP packets it gets from the servers, then does a reverse lookup
to translate them back to names -- unless you use "ntptrace -n".

> If I query the time sources directly, same error:
> 
> # ntptrace sputnik-1.ca.nortel.com
> zcarh00x.ca.nortel.com: timed out, nothing received
> ***Request timed out
> 
> # ntptrace sputnik-2.us.nortel.com
> zsclyntp.us.nortel.com: timed out, nothing received
> ***Request timed out
> 
> # ntptrace sputnik-3.europe.nortel.com
> sputnik-3.europe.nortel.com: stratum 1, offset 0.000007, root distance 
> 0.000000, refid 'GPS'

What happens if you try "ntpq -n -c rv sputnik-1.ca.nortel.com" ?

> I've confirmed all the time sources are working properly, in fact the 
> clinets have no problem syncing with any of them.
> # ntpq -c pe
>      remote           refid      st t when poll reach   delay   offset 
> jitter
> ==============================================================================
> +zcarh00x        .GPS.            1 u   49  256   17    1.906   -0.390 
> 0.278
> *zsclyntp        .GPS.            1 u   43   64   37  103.143   -0.719 
> 0.535
> +whars09p        .GPS.            1 u   45  256   17   95.408   -0.084 
> 0.657
> 
> 
> Its not a routing or DNS mis-config.  I've confirmed NTPTRACE 4.1.1 works 
> properly with same ntp.conf.
> 
> Has anyone else seen this problem?  Suggested solutions?

ntptrace in 4.1.1 is a C program that makes NTP time requests.
ntptrace in 4.2.0 is a Perl script that uses ntpq to read information.
I'd guess your Spectracoms are configured with "restrict default noquery".

-- 
                      Ronan Flood <R.Flood at noc.ulcc.ac.uk>
                        working for but not speaking for
             Network Services, University of London Computer Centre
     (which means: don't bother ULCC if I've said something you don't like)




More information about the questions mailing list