[ntp:questions] Help appreciated

Richard B. Gilbert rgilbert88 at comcast.net
Tue May 12 20:14:28 UTC 2009


Towli wrote:
> Hi all
> 
> I found out that our Inet-Provider has stratum 1 servers available, and 
> i intended to use these together with our newly acquired DCF77 machinery.
> I configured our DFC77 and the 2 external stratum 1 servers on our 2 
> core switches to deliver Ntp to the rest of the network...
> Unfortunately (?) my switches seem to prefer these external time sources 
> as opposed to our own internal DCF77 Furthermore i can se (from output 
> below) that our internal DCF77 is listed as "invalid" ? ? ? Why is that?
> I have not yet set up our backup DCF77. When i do - should i loose the 2 
> external sources and use solely internal NTPservers or what would your 
> recommendations  / best practice be, based on the output below ?
> 
> Any advice is highly appreciated as i am on thin ice here, and it is an 
> important task to implement NTP correctly on out network.
> 
> Output: (for "anonymous" purposes in regards to my job i have marked the 
> external ip's with xxx)
> 
> 
> /Towli - Denmark
> 
> 
> #sh ntp ass de
> 193.xxx.xxx.xxx configured, our_master, sane, valid, stratum 1
> ref ID .GPS., time CDB44C38.435EEBEB (21:27:52.263 CEST Tue May 12 2009)
> our mode client, peer mode server, our poll intvl 64, peer poll intvl 64
> root delay 0.00 msec, root disp 1.69, reach 377, sync dist 9.247
> delay 6.13 msec, offset 4.1194 msec, dispersion 4.30
> precision 2**18, version 3
> org time CDB44C68.7CC60B4F (21:28:40.487 CEST Tue May 12 2009)
> rcv time CDB44C68.7C8118AB (21:28:40.486 CEST Tue May 12 2009)
> xmt time CDB44C68.7AEB32B4 (21:28:40.480 CEST Tue May 12 2009)
> filtdelay =     6.13    6.12    5.57    8.21    5.10    6.33    5.98    
> 5.05
> filtoffset =    4.12    6.02    8.18    9.58   13.80   16.39   19.32   
> 22.59
> filterror =     0.02    0.99    1.97    2.94    3.92    4.90    6.64    
> 8.59
> 
> 193.zzz.zzz.zzz configured, selected, sane, valid, stratum 1
> ref ID .GPS., time CDB44C3C.894CBE10 (21:27:56.536 CEST Tue May 12 2009)
> our mode client, peer mode server, our poll intvl 64, peer poll intvl 64
> root delay 0.00 msec, root disp 1.80, reach 377, sync dist 7.889
> delay 5.43 msec, offset 4.2101 msec, dispersion 3.37
> precision 2**18, version 3
> org time CDB44C74.7DD78E42 (21:28:52.491 CEST Tue May 12 2009)
> rcv time CDB44C74.7D760EA9 (21:28:52.490 CEST Tue May 12 2009)
> xmt time CDB44C74.7C0F49F3 (21:28:52.484 CEST Tue May 12 2009)
> filtdelay =     5.43    6.73    5.94    5.89    6.13    4.82    5.07    
> 5.49
> filtoffset =    4.21    5.25    7.64   10.08   12.67   16.30   17.74   
> 20.33
> filterror =     0.02    0.99    1.97    2.94    3.92    4.90    5.87    
> 7.83
> 
> 10.200.138.20 configured, sane, invalid, stratum 1
> ref ID .DCF., time CDB44934.00000000 (21:15:00.000 CEST Tue May 12 2009)
> our mode client, peer mode server, our poll intvl 1024, peer poll intvl 
> 1024
> root delay 0.00 msec, root disp 0.00, reach 377, sync dist 21.103
> delay 3.27 msec, offset -11.4426 msec, dispersion 7.02
> precision 2**6, version 3 <------------------------- what does this 
> precission indicate ? Better or worse than 2**18 ?

Precision is normally expressed as a *negative* power of two!  Assuming 
that somebody dropped the minus signs somewhere, 2**-18 is better than 
2**-8.  Precision is the smallest possible difference between two 
successive readings of the clock.




More information about the questions mailing list