[ntp:questions] NTP time synchronization issue in linux embedded target

David Woolley david at ex.djwhome.demon.invalid
Wed Sep 3 07:12:11 UTC 2014


On 02/09/14 19:43, Murugesh S wrote:
> 1. Why and what causes the drift to reach 500.00 on certain systems

Normally broken hardware.

> # Drift file
> driftfile /etc2/ntp.drift
>
> restrict 127.0.0.1
>

There are no server lines; I don't understand how you can have a server 
association.

However, in this case, the server is not a valid NTP server.

> org time 00000000.00000000 (Feb 07 00:28:16 CST 2036)

This should be your current time as of when you sent the poll.

> rcv time d794b214.1bc65df9 (Aug 12 10:32:36 CDT 2014)

This is where you are getting the reference time from and it should be 
updating.

> xmt time 00000000.00000000 (Feb 07 00:28:16 CST 2036)

This should be very close to the rcv time.

> filtdelay =  55.87  55.70  55.68  55.61  115.01  56.03  55.72  59.62
> filtoffset =  -881879.45  -881396.98  -880822.61  -880219.19  -879752.22  -879248.85  -878747.85  -8

Although not a cause of your problem:

> precision 2**-6,

is pretty poor.   It is basic MSDOS precision.  You should expect 2**-10 
for an oldish NTP capable Unix based server, and the high teens, in the 
exponent, for a modern server.



More information about the questions mailing list