This question is pretty specific, but I hope someone will have seen this error elsewhere. I am configuring a machine running OS X 10.5.8 to be an NTP server for machines connected to a LAN that is not connected to the Internet. I am not too worried about knowing the "right" time on all the machines, but rather worried about making sure everyone has the same notion of time.
I configured the NTP daemon on Mac by turning on the Set date and time automatically in System Preferences, using the server's clock, 127.127.1.0 as the reference clock. I figured I should see if the server can NTP query itself before proceeding to the clients.
The weird part is when I run the ntpq -p command in a command-prompt when connected to my D-Link DIR-655 (firmware: 1.33), it hangs for about a minute or so each time before finally giving me some output. I thought the problem might have to do with Port Forwarding, so I configured the router to forward port 123 for the IP of the server, but that did not improve the situation.
When I run the ntpq -p command on my school's network, on a Linksys WRT54G router, or with the wireless Airport card turned off - I have absolutely no problems - the command returns a response instantly.
Is this normal? I can see why a query might take a minute or so, but I don't understand why one router does it faster than the other.
I tried messing around with the ntp.conf file adding the burst, minpoll, and maxpoll options:
server 127.127.1.0 burst minpoll 4 maxpoll 5
Figuring that perhaps I am polling too often and the configuration file is slowing me down, but even with this, the ntpq still hangs on the D-Link DIR-655, but does just fine on the other routers.
Any thoughts on where the lag is coming from or if the lag is even a problem?