[ntp:questions] ntpd sometimes goes wacko, all peers in INIT, packets mostly ignored, but restart fixes it

David Woolley david at ex.djwhome.demon.co.uk.invalid
Sun Oct 7 13:03:27 UTC 2007


In article <4703FDFC.2030100 at math.arizona.edu>,
aprl at math.arizona.edu (Alexander Perlis) wrote:

 > You haven't really provide any hard information, but a wild guess
 > would be misuse of local clock in an isolated time island.

> Thank you for trying to help us. My apologies as a newbie as far as 

As a newbie, you should not be using advanced restrict options, as
a lot of failures are the result of using too powerful restrictions.

>    1 39197  8000   yes   yes  none    reject

> assID=39197 status=8000 unreach, conf, no events,

> What's perhaps strange here is that it says "unreach" and has 
> "dstadr=0.0.0.0". The non-wacko ntpds have "reach" and 

I was expecting a reason, not unreachable.  In any case, the associations
list indicates that it is reject not unreachable.

However:

>            Originator Timestamp: 3400431260.975470633 (2007/10/03 13:14:20)
>            Receive Timestamp:    3400432606.929923057 (2007/10/03 13:36:46)
>            Transmit Timestamp:   3400432606.930152893 (2007/10/03 13:36:46)
>              Originator - Receive Timestamp:  +1345.954452395
>              Originator - Transmit Timestamp: +1345.954682230

These are as good as drop dead conditions.  If your client were to accept the
time from this server, the ntpd daemon process would suicide, because the 
time discrpancy was too large to trust and either it, or all its servers, would
be presumed broken, but it doesn't know which.

PS Your email program appears to be broken; your reply to the mail to news
gateway seems to be missing an In-Reply-To header, and, as a result, the
threading has been broken.  I've added back the thread root message ID 
to try to reconnect it, but it may not work.  Accessing the newsgroup
directly would be better.

quit




More information about the questions mailing list