time/ntp[d]

[michael]

On Wed, 2008-08-06 at 18:16 +0200, Björn Persson wrote:
> michael wrote:
> > mkb at veri:~$ sudo /usr/sbin/ntpq -p
> >      remote           refid      st t when poll reach   delay   offset
> > jitter
> > ===========================================================================
> >===
> >  utserv.mcc.ac.u 193.62.22.98     2 u    7   64    1    2.575  348421. 
> > 0.001
> >
> > but not sure what that means....
> 
> · Your time is not synchronized yet, as the line begins with a space. If NTPD 
> thinks the server seems reliable it will synchronize to it, and an asterisk 
> should appear before the server's name.
> · remote: NTPD is talking to utserv.mcc.ac.u... (The name is truncated.)
> · refid: utserv is synchronized to 193.62.22.98.
> · st: utserv is at stratum 2, so you'll be at stratum 3.
> · I'm not sure about the "t" field. It might mean type of timesource.
> · when: NTPD latest queried utserv 7 seconds ago.
> · poll: The interval between queries is currently 64 seconds. This will 
> gradually increase to 1024 seconds if the communication works well.
> · reach appears to be an array of eight bits represented as an octal number. 
> It keeps track of whether responses were received to the eight latest 
> queries. As you just started NTPD there's only been one query so far. If no 
> packets are lost the value will eventually be 377.
> · delay: A roundtrip to utserv and back takes 2.575 milliseconds. I don't know 
> if this is the latest roundtrip or some sort of average.
> · offset: Your time differs with 348.421 seconds from utserv's time. When 
> synchronized, NTPD should be able to keep this below a tenth of a second.
> · jitter is, as far as I understand, a measure of the server's stability. If 
> its time doesn't appear to be running uniformly, the jitter value will be 
> high. (Varying delays in communication can cause this.)
> 
> Björn Persson

most useful! thanks!!

M