G'day, I have an ntp server or two running on hosts using ip aliases dedicated to the ntp service. The problem is the ntpd is not responding on the same interface/alias that it recieves queries on. This means clients are not happy;
$ sudo ntpdate -d -b -s ntp.melb.apana.org.au transmit(203.12.236.41) receive(203.12.236.20) receive: server not found transmit(203.12.236.41) receive(203.12.236.20) receive: server not found note that the 203.12.236.41 address is eth0:1, and 203.12.236.20 is eth0 on the same host. Is it a reportable bug that ntpd does not respond from the interface the query was recieved on? Is there any way to get ntpd to bind to a particular interface/alias, or the specify it's outgoing address? If not, any hints as to how to resolve this? I don't want to make the ntp.melb.apana.org.au IP address eth0 if I can avoid it. I have a feeling I might be able to do some scary iproute2 stuff to route outgoing udp from the ntp port via the ntp.mel.apana.org.au alias... but there must be an easier way. -- ---------------------------------------------------------------------- ABO: finger [EMAIL PROTECTED] for more info, including pgp key ---------------------------------------------------------------------- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
