[ntp:questions] Re: NTP, Mac OS X & Cisco 837 - some stats:

Richard B. Gilbert rgilbert88 at comcast.net
Wed Apr 20 12:52:16 UTC 2005


viz wrote:

>Some interesting stats come up with ntpq -p:
>
> ntpq -p
>     remote           refid      st t when poll reach   delay   offset
>jitter
>============================================================================
>==
>+bm1.asia.apple. time1.euro.appl  2 u  118  128  377  127.025  -227.69
>401.635
>+mel001.pacific. murgon.cs.mu.oz  2 u  101  128  377   36.358  -219.35
>435.265
>+marvin.ci.com.a .PPS.            1 u   63  128  377   36.375  -246.40
>40.575
>*murgon.cs.mu.oz .GPS.            1 u  114  128  377   32.412  -231.68
>410.731
>+ntp1.cs.mu.oz.a .GPS.            1 u  114  128  377   42.499  -228.29
>418.720
>-bob.greenwood.w ntp.hobart.nml.  2 u  124  128  377  206.531  -169.20
>474.324
>-ntp1.tpgi.com.a murgon.cs.mu.oz  2 u  120  128  377   21.104  -285.12
>415.440
>+levy.catalyst.n gen1.ihug.co.nz  3 u  120  128  377   62.818  -226.84
>414.083
>
>
>ntpq -p
>     remote           refid      st t when poll reach   delay   offset
>jitter
>============================================================================
>==
>-bm1.asia.apple. time2.euro.appl  2 u  92m  256  377  129.815   -8.807
>8.620
>-mel001.pacific. murgon.cs.mu.oz  2 u  92m  256  377   38.591    0.902
>0.220
>-marvin.ci.com.a .PPS.            1 u  93m  256  377   87.344  -41.230
>32.827
>*murgon.cs.mu.oz .GPS.            1 u  92m  256  377   33.880  -13.720
>2.081
>+ntp1.cs.mu.oz.a .GPS.            1 u  92m  256  377   34.493  -13.122
>4.711
>+bob.greenwood.w ntp.hobart.nml.  2 u  92m  256  377   87.686  -13.215
>1.836
>-ntp1.tpgi.com.a murgon.cs.mu.oz  2 u  92m  256  367   26.237  -68.400
>1.196
>+levy.catalyst.n gen3.ihug.co.nz  3 u  93m  256  377   57.825  -14.396
>3.948
>
> 
>
>ntpq -p
>     remote           refid      st t when poll reach   delay   offset
>jitter
>============================================================================
>==
>-bm1.asia.apple. time1.euro.appl  2 u 107m  128   36  129.001  -18.558
>4.063
>+mel001.pacific. murgon.cs.mu.oz  2 u 105m  128   37   43.009  -24.018
>10.079
>-marvin.ci.com.a .PPS.            1 u 104m  128   37   63.324  -41.983
>161.509
>+murgon.cs.mu.oz .GPS.            1 u 106m  128   37   35.328  -24.333
>4.102
>*ntp1.cs.mu.oz.a .GPS.            1 u 106m  128   37   44.048  -21.404
>0.326
>-bob.greenwood.w ntp.hobart.nml.  2 u 105m  128   37   95.627  -31.444
>12.165
>xntp1.tpgi.com.a ntp1.tip.csiro.  2 u 104m  128   37   23.700  -81.640
>10.341
>-levy.catalyst.n msltime.irl.cri  2 u 105m  128   37   56.696  -31.656
>10.730
>
>The jitter numbers have a huge range - any ideas why?
>
>/viz
>
>  
>
I suspect it might have something to do with your poll intervals!  Those 
servers were last polled 105 minutes ago!  It looks as if your minpoll 
is still 12 and maxpoll is still 17.  NTP can't work very well with such 
a misconfiguration!  Minpoll should be six and maxpoll should be ten.  
These are the default values if they are not specified.

I believe another post in this thread mentioned a template configuration 
file in /etc/private (?) that was copied over the /etc/ntp.conf file 
every time ntpd was started.  Fix that file or fix /etc/ntp.conf and 
start ntpd by hand!!!!   I believe that, having done so, you will see 
much quicker synchronization and much tighter (smaller offsets) 
synchronization.  Your poll intervals should then start at 64 seconds 
and ramp up to 1024 seconds as ntpd locks in.



More information about the questions mailing list