On Tuesday 30 October 2007 02.43:35 Darrin Chandler wrote: > After initially settling down, the bit of scribbling seems due to > zeroing in on the clock drift, which it just about nailed at this time.
While the delay certainly plays its part, I'd still say openntp performs
abysmally bad compared to ntpd, the jitter is extremely high, and the
offset is also the highest of all configured servers:
remote refid st t when poll reach delay offset jitter
==============================================================================
-193.247.72.14 130.149.17.21 2 u 522 1024 377 4.004 2.070 0.058
+195.216.64.208 129.69.1.153 2 u 633 1024 377 3.150 -1.263 0.171
217.147.223.78 129.132.2.21 3 u 860 1024 377 3.189 0.336 0.549
+194.97.156.5 131.188.3.222 2 u 758 1024 377 18.678 -0.006 0.272
*193.192.51.156 193.62.22.74 2 u 813 1024 377 19.937 -1.666 0.026
-193.228.143.12 192.36.144.23 2 u 532 1024 377 38.144 3.415 0.578
193.138.215.196 129.132.2.21 3 u 503 1024 377 1.841 1.401 0.205
212.103.65.133 129.132.2.21 3 u 530 1024 377 1.372 0.854 0.257
-162.23.41.34 162.23.3.171 3 u 547 1024 377 3.352 3.557 0.378
-129.132.2.21 129.132.2.22 2 u 925 1024 377 3.452 0.575 0.356
-140.99.51.115 131.216.1.101 3 u 645 1024 377 164.743 -5.634 12.768
Can somebody closer to that server configure it to eliminate the bad
network?
I still strongly recommend that people use only proper ntp implementations
(which implement the full algorithms as per the RFC) as public timeservers.
As far as I know ntpd is the only such implementation.
The network protocol is only the smallest part of NTP. The big part is in
the logic, and afaict openntp doesn't implement most of this.
cheers
-- vbi
--
We believe in the immeasurable societal good that stems from the unequal
treatment of gays, in much the same way that we believe in Intelligent
Design, cold fusion and The SCO Group’s ability to make money off
something other than patent litigation.
-- Brandon Burt, Salt Lake City Weekly, July 21, 2005
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ timekeepers mailing list [email protected] https://fortytwo.ch/mailman/cgi-bin/listinfo/timekeepers
