[ntp:questions] Re: How long do I have to wait for sync?

Keerati Inochanon keerati.ai at gmail.com
Wed Jun 1 19:24:51 UTC 2005


"David J Taylor" <david-taylor at blueyonder.co.not-this-bit.nor-this-part.uk> 
wrote in message news:cnkne.43397$G8.22777 at text.news.blueyonder.co.uk...
> Keerati Inochanon wrote:
>> "David J Taylor"
>> <david-taylor at blueyonder.co.not-this-bit.nor-this-part.uk> wrote in
>> message news:GAdne.43144$G8.6443 at text.news.blueyonder.co.uk...
>>> Keerati Inochanon wrote:
>>> []
>>>> ntpq -p on the slaves shows:
>>>> remote        refid        st t when poll reach    delay    offset
>>>> jitter ===============================================
>>>> master.name    .INIT.      16 u  -      64    0         0.000  0.000 
>>>> 4000.00
>>>
>>> The "0" value for reach says the slave cannot connect to the master.
>>> As a starter, try the command:
>>>
>>>  ntpq -p master
>>>
>>> from one of the slaves to see if there is a connectivity issue.
>>
>> Doing the command
>>
>> ntpq -p master
>>
>> on the slave works just fine. It prints out exactly the same thing as
>> doing the command on the master machine. However, doing ntpq -p on
>> the slave still give the same result with the refid of the master
>> being .INIT. and in stratum 16.
>>
>> Any thoughts?
>
> Well, I have seen something similar on the systems here (Windows XP, 
> Window 2000) when they have intermittent contact with some servers outside 
> the LAN, but good connectivity with servers inside the LAN.  This seems to 
> be an intermittent problem, though, and it's not /all/ servers outside the 
> LAN which get disconnected.  Even with the machines on continuously, 
> sometimes it's OK one day and faulty the next.  Never been able to pin 
> that one down.

Thank you very much for your response. The setup I have is also Windows XP + 
Windows 2000 machines. Maybe there is some connectivity problems there?

The problem used to be intermittent. The slave used to be able to connect to 
the master after about 20 minutes or so, but this time it does not, even 
when I left both of them running over night. I am able to ping both the 
machines, and the firewall ports for the machines are also open.

>
> Next question - any restrict or notrust or similar lines in your 
> ntpd.conf?  You didn't show any, so perhaps there aren't any!
>

I do not have any restrict or notrust line in the ntpd.conf. The one I 
posted here is pretty much what I have look like.

> Is the IP address of the master correct?

Yes.. actually... no..
I was about to say yes.. then I decided to check.. and then the answer is 
no. The IP address much have changed by dhcp. I have always assumed that it 
did not change as the ntpq -p always show the correct reverse resolved name 
of the ip. I guess it must be in the cache.. (now I feel dumb).

> Perhaps you are expecting too much?  It could take a couple of minutes for 
> the master to get into the synched state (shown by the "*" against one of 
> the master's servers), and only after that will the slaves start to see 
> the master as a suitable source of time.  (I think this is the case, 
> others will correct me).

> The other thought is that NTP is best suited to systems which are left on 
> all the time, so at least the server connected to the internet should run 
> continously.  Having said that, NTP works just fine on my portable, which 
> is switched off most of the time.
>
> Sorry not to help more...

You actually helped with the sanity check =) Thank you very much! I 
appreciate it!

Best regards,
Keerati 





More information about the questions mailing list