[pkg-ntp-maintainers] Bug#511931: Bug#511931: NTP servers (Debian Pool) reject connections

[Kurt Roeckx]

On Thu, Jan 15, 2009 at 07:13:25PM +0100, Michelle Konzack wrote:
> Package: ntp
> Version: 1:4.2.2.p4+dfsg-2
> Severity: normal
> 
> Hello Maintainer,
> 
> I was wondering about the time difference in my local network  which  is
> normaly synced with my local NTP server...  It seems, my Intranet server
> has problems connection to the 4 Debian Pool Servers.  So I have tested
> it from my Laptop with:
> 
> fnc() { for IP in 212.201.68.154 213.239.197.201 81.169.180.21 84.200.248.203 131.234.137.24 141.40.103.101 217.160.215.119 78.46.108.116 88.198.39.175 89.110.145.252 141.40.103.103 78.47.136.197 81.169.180.26 89.110.145.252 131.234.137.23 213.198.65.20 80.190.97.205 89.149.209.134 131.234.137.23 141.40.103.103 ; do DATE=$(date --rfc-822) ; echo "Check NTP server on:  ${IP}  [${DATE}]" ; rdate ${IP} ; echo ; done ; } ; fnc >/root/time_server.log 2>&1
> 

"default rdate uses the RFC 868 TCP protocol.", which is not (s)ntp.  I do
not expect those servers to run a time server (on port 37).  They do
answer to (s)ntp packets (udp port 123).

rdate has a "-n" option to use sntp.

I understand that it's your server that is having a problem?
What does "ntpq -c peers" return?  Did the log file show anything,
specialy at the time it got started?

Does restarting ntp fix the problem?



Kurt