[ntp:questions] Re: NTP Stats within norm?

Brian Utterback brian.utterback at sun.removeme.com
Thu May 25 19:38:11 UTC 2006


Richard B. Gilbert wrote:
> Scott L. Glidden wrote:
>> Hope you don't mind a top-post...
>>
>>      remote           refid      st t when poll reach   delay   offset 
>> jitter
>> ============================================================================== 
>>
>> -time.nist.gov   .ACTS.           1 u   87  512   37   70.905   10.262 
>> 1.691
>> -ntp-nasa.arc.na .GPS.            1 u   94  512  377  102.491   10.018 
>> 1.620
>> -time-a.nist.gov .ACTS.           1 u  233  512  377   45.727    2.009 
>> 4.349
>> -time-nw.nist.go .ACTS.           1 u  320  512  377   87.205   13.117 
>> 0.393
>> -clock.isc.org   .GPSi.           1 u  264  512  377  107.590   14.738 
>> 1.249
>> *ntp1.twc.weathe .CDMA.           1 u  334  512  377   34.050    5.457 
>> 1.018
>> -ntp2.twc.weathe .CDMA.           1 u  262  512  377   34.031    6.166 
>> 1.212
>> -ns1.usg.edu     130.207.244.240  2 u  244  512  377   26.068    7.831 
>> 1.776
>> +ns2.usg.edu     .GPS.            1 u  346  512  377   31.300    4.523 
>> 0.957
>>  ns3.usg.edu     65.212.71.102    2 u  255  512  377   26.609    7.583 
>> 1.248
>> -ns4.usg.edu     130.207.244.240  2 u  233  512  377   31.934    6.318 
>> 3.012
>> -caesar.cs.wisc. 128.105.201.11   2 u  332  512  377   42.701    8.937 
>> 0.746
>>  now.cis.okstate .STEP.          16 u  12h 1024    0    0.000    0.000 
>> 0.000
>> -ntp2.usno.navy. .USNO.           1 u  213  512  377   54.995    2.985 
>> 4.446
>> +tock.usno.navy. .USNO.           1 u  218  512  377   56.119    4.484 
>> 0.885
>> -elastic.org     204.152.184.72   2 u  231  512  377  116.555    6.842 
>> 0.172
>> -216.106.191.180 192.5.41.41      2 u  241  512  377   69.417    7.609 
>> 1.177
>> -ns1.it2.net     204.123.2.72     2 u  317  512  377   89.060    9.447 
>> 1.447
>> -opium.apexvoice 138.23.180.126   3 u  333  512  377   89.006    9.750 
>> 1.280
>>
>>
>>> What does the output of an "ntpq -p" show?
>>>
>>> David
>>>
>>
>>
>>
> 
> Lose all the servers with delays greater than 40 milliseconds.  The 
> error in transmitting time from server to client is bounded by one half 
> the round trip delay; it may be better but it can't get worse than that. 
>  Seven servers should be sufficient for the utmost in reliability.  Ten 
> is the maximum that ntpd will include in the "selection set"
> 

And yet the server with the best jitter is also the one with the
largest delay. And it shows an offset that is right in line with
a group of servers that is below 40 in delay.

Not to muddy the waters, but the shortest delay may tend to be
a better choice, but not always.

-- 
blu

Rose are #FF0000, Violets are #0000FF. All my base are belong to you.
----------------------------------------------------------------------
Brian Utterback - OP/N1 RPE, Sun Microsystems, Inc.
Ph:877-259-7345, Em:brian.utterback-at-ess-you-enn-dot-kom




More information about the questions mailing list