🚀 go-pugleaf

I've just started using NTP revision 3.4.1.9, patchlevel 13, from
UUNET on our Sun 4 (OS 4.1) systems and have what appears to be a bit
of a problem.

Here is the output from an ntpdc request made of one of our Sun
systems running ntp:

% ntpdc mozart
   (rem)  Address   (lcl)      Strat Poll Reach    Delay   Offset    Disp
==========================================================================
+ravel           192.5.98.5        2   64  377      20.0    425.0   3529.0
+asihub          192.5.98.5        2   64  377      20.0    402.0   3506.0
*handel          192.5.98.5        2   64  377      20.0    376.0   3504.0


Here is the output from an ntpdc request made of one of our MIPS
systems running ntp:

% ntpdc cedar
   (rem)  Address   (lcl)      Strat Poll Reach    Delay   Offset    Disp
==========================================================================
+ravel           192.5.98.41       2  128  377      20.0    -19.0      9.0
+asihub          192.5.98.41       2  256  377      20.0    -40.0      7.0
*handel          192.5.98.41       2   64  377      20.0    -67.0      7.0

The three servers listed are all ULTRIX systems, some VAX's, some RISC.


What concerns me is that the 'Offset' and 'Disp' values for the Sun
system seem to be quite high.  I've reviewed the numbers of a period
of days and these numbers are always much higher on the Sun systems
than on the MIPS or DEC systems.  The values of 'Disp' will vary from
500 to 31995 and the offset from 400 to 1000.

Anyone have any idea why this might be happening?  The Sun systems
are all Sparc 1+'s running OS 4.1, the DEC's are running UWS 4.0, the
MIPS are running 4.51.

The sun systems have no peer entries, just three server entries.

Thanks!


-- 
Dan O'Neill     Cadence Design Systems, San Diego, CA
UUCP: {uunet|ncr-sd}!asihub!dano     ...cadence!dano
Domain: dano@autosys.com     dano@cadence.com