Hi all,

    Recently I made some modifications in the code of NTP just to log the r=
equests that arrive on the server. In this logs I put date, time, IP addres=
s of client and server, dootdispersion and rootdelay of the client (taken f=
rom the 'pkt' variable).
    All that works fine, but I want to know why of the discrepancy in the v=
alues of rootdispersion. Take a look on the logs:
---------
date; time; rootdispersion(ms); rootdelay(ms); client; server;

11/08/2004; 16:56:01; 216.461; 0.229; 128.0.0.213; 128.0.0.216;
11/08/2004; 16:56:03; 216.492; 0.229; 128.0.0.213; 128.0.0.216;
11/08/2004; 16:56:05; 216.522; 0.229; 128.0.0.213; 128.0.0.216;
11/08/2004; 16:56:07; 216.553; 0.229; 128.0.0.213; 128.0.0.216;
11/08/2004; 16:56:09; 958.069; 0.244; 128.0.0.213; 128.0.0.216;
11/08/2004; 16:56:11; 958.099; 0.244; 128.0.0.213; 128.0.0.216;
11/08/2004; 16:56:13; 207.321; 0.244; 128.0.0.213; 128.0.0.216;
11/08/2004; 16:56:15; 207.352; 0.244; 128.0.0.213; 128.0.0.216;
---------

        The proximity for the requests is because of the iburst. I just don=
=B4t understand why at 16:56:07 the dispersion if 216.553ms and at 16:56:09=
 this value jumps to 958.069ms, going back to 207.321ms at 16:56:13. Is tha=
t normal?
Thanks for any help,

V=EDtor

Vitor,

Variations like this are not unusual. The root values are summed on the 
path to the primary server and can change when a server along the path 
switches to a different server or if a packet is lost, etc.

Dave

V�tor Claudino dos Santos wrote:

> Hi all,
> 
>     Recently I made some modifications in the code of NTP just to log the requests that arrive on the server. In this logs I put date, time, IP address of client and server, dootdispersion and rootdelay of the client (taken from the 'pkt' variable).
>     All that works fine, but I want to know why of the discrepancy in the values of rootdispersion. Take a look on the logs:
> ---------
> date; time; rootdispersion(ms); rootdelay(ms); client; server;
> 
> 11/08/2004; 16:56:01; 216.461; 0.229; 128.0.0.213; 128.0.0.216;
> 11/08/2004; 16:56:03; 216.492; 0.229; 128.0.0.213; 128.0.0.216;
> 11/08/2004; 16:56:05; 216.522; 0.229; 128.0.0.213; 128.0.0.216;
> 11/08/2004; 16:56:07; 216.553; 0.229; 128.0.0.213; 128.0.0.216;
> 11/08/2004; 16:56:09; 958.069; 0.244; 128.0.0.213; 128.0.0.216;
> 11/08/2004; 16:56:11; 958.099; 0.244; 128.0.0.213; 128.0.0.216;
> 11/08/2004; 16:56:13; 207.321; 0.244; 128.0.0.213; 128.0.0.216;
> 11/08/2004; 16:56:15; 207.352; 0.244; 128.0.0.213; 128.0.0.216;
> ---------
> 
>         The proximity for the requests is because of the iburst. I just don�t understand why at 16:56:07 the dispersion if 216.553ms and at 16:56:09 this value jumps to 958.069ms, going back to 207.321ms at 16:56:13. Is that normal?
> Thanks for any help,
> 
> V�tor