Hi Kurt, Kurt Roeckx schrieb: > On Mon, Oct 16, 2006 at 06:28:17PM +0200, Christoph Martin wrote: >> I updated one some of my systems ntp from sarge to current testing. >> On one of the systems ntpd is not adjusting the time after the inititial >> sync. >> >> ntpq shows an every increasing time difference: >> >> ntpq -n -c peers >> remote refid st t when poll reach delay offset >> jitter >> >> ============================================================================== >> 134.93.226.19 134.93.227.254 3 u 18 64 377 0.524 >> -14813. 1689.13 >> 134.93.227.254 131.188.3.220 2 u 39 64 377 0.443 >> -15216. 2153.18 >> 134.93.226.2 134.93.225.254 5 u 34 64 377 0.430 >> -15255. 2167.75 >> >> After some days it is over an hour and I have to restart ntpd to get a >> sync once again. Only after every restart I get a sync. > > Does /var/log/daemon.log contain anything useful? It says that it > synched once?
Oct 16 17:50:10 gonzo ntpd[26207]: ntpd exiting on signal 15
Oct 16 17:50:35 gonzo ntpd[27045]: ntpd [EMAIL PROTECTED] Mon Oct 2
17:50:06 UTC 2006 (1)
Oct 16 17:50:35 gonzo ntpd[27046]: precision = 33.000 usec
Oct 16 17:50:35 gonzo ntpd[27046]: Listening on interface wildcard,
0.0.0.0#123 Disabled
Oct 16 17:50:35 gonzo ntpd[27046]: Listening on interface wildcard,
::#123 Disabled
Oct 16 17:50:35 gonzo ntpd[27046]: Listening on interface lo, ::1#123
Enabled
Oct 16 17:50:35 gonzo ntpd[27046]: Listening on interface eth0,
fe80::20e:a6ff:fe54:29c2#123 Enabled
Oct 16 17:50:35 gonzo ntpd[27046]: Listening on interface lo,
127.0.0.1#123 Enabled
Oct 16 17:50:35 gonzo ntpd[27046]: Listening on interface eth0,
10.100.64.1#123 Enabled
Oct 16 17:50:35 gonzo ntpd[27046]: kernel time sync status 0040
Oct 16 17:50:35 gonzo ntpd[27046]: frequency initialized 0.000 PPM from
/var/lib/ntp/ntp.drift
Oct 16 17:50:45 gonzo ntpd[27046]: synchronized to 134.93.227.254, stratum 2
Oct 16 17:48:59 gonzo ntpd[27046]: time reset -105.185036 s
Oct 16 17:48:59 gonzo ntpd[27046]: kernel time sync disabled 0001
> I'm guessing it says it synched to 134.93.227.254 in that
> case.
Correct.
> Are you using the same config on all hosts?
Yes. Basically I use my own round robin dns for the ntp servers, which
result in the same servers in the end.
> The hosts seem to have a rather large difference in time between them.
> It says:
> -14813.
> -15216.
> -15255.
>
> Which is 442 ms difference between the clocks. I think ntp rejects all
> clocks because they don't agree on the time. I think the algorithm
> requires 3 clocks to agree before it accepts them.
Well, I am not shure about that because the other host which works has
the following stats:
134.93.226.2 134.93.225.254 5 u 27 64 377 0.262 -3116.1
1075.62
134.93.226.19 134.93.227.254 3 u 54 64 377 0.330 -1174.2
1370.48
*134.93.227.254 131.188.3.220 2 u 48 64 377 0.313 -1791.6
917.537
Here I have a difference of about 2000! and its working.
> I also find it weird that the first (134.93.226.19) which says it's
> synched to the second (134.93.227.254) has such a big difference in
> time. They all seem to be in the same network, so they really shouldn't
> be different more than a few ms.
That is true. I have to look into it. But it might be because the two
servers have the same list of upstream ntp servers but decide to sync to
different ones. But I should make shure that they have the same time. Do
you have any idea how to bind two ntp daemons tightly?
>> reference ID: [83.84.69.80]
>
> Where does that 83.84.69.80 come from? I didn't see that in your list
> of peers?
I did some search on google. This is a bug in ntpd or ntpdc. The values
are ascii codes for STEP.
The weird thing is on the host that I don't get any error message why it
can't sync to any server.
Christoph
--
============================================================================
Christoph Martin, Leiter der EDV der Verwaltung, Uni-Mainz, Germany
Internet-Mail: [EMAIL PROTECTED]
Telefon: +49-6131-3926337
Fax: +49-6131-3922856
signature.asc
Description: OpenPGP digital signature
