ntpd woes

Top Page
Attachments:
Message as email
+ (text/plain)
Delete this message
Reply to this message
Author: David A. Sinck
Date:  
Subject: ntpd woes
I'm attempting to beat down an ntpd problem and rtfming hasn't helped
so far. I've got a central server that requests information from
public time servers and happily alternates normally between the
'presently best' servers. I've got several 'slaves' that talk to the
central server and *should* get sync'd to that.

However, none of the slaves want to do that. They see the master just
fine, they talk to it fine, but they just don't like it. Even when
it's the only option. I've forced the date with ntpdate and it gets
that from the server no problem. (Unless it did a really good job of
faking talking to the central server.)

The master file looks like:
+-------------
| restrict default ignore
| restrict 127.0.0.1 
| restrict colo.subnet mask 255.255.255.240 notrust nomodify notrap
| restrict colo2.subnet mask 255.255.255.240 notrust nomodify notrap
| server x.x.x.x
| server x.x.x.x
| server x.x.x.x
| server x.x.x.x
| server  127.127.1.0     # local clock
| fudge   127.127.1.0 stratum 10  
| driftfile /etc/ntp/drift
| broadcastdelay  0.008

+-------------

The simplest ntp.conf file looks like:
+-------------
| server central.server.ip
| driftfile /etc/ntp/drift
| broadcastdelay 0.008

+-------------

Another 'slave' has the local clock conf lines as fallbacks and ALWAYS
picks that.

Pointers?

TIA,

David