NTP time source behaviour

S

stan drews

Hi

I have posted twice before with regard to NTP config and settings and am very grateful for all the responses I have received. However I am still pulling my hair out with regard to a specific item. I hope someone can take the time to read this and offer some assistance. I have been unable to find any documentation/articles anywhere that detail this particular issue.

Scenario: 1 x server 2016 domain, 2 x hardware clocks. PDCe configured with IP addresses of both hardware clocks.

Issue: When issuing w32tm /query /source on the PDCe the returned result is one of the hardware clock IP addresses. This is as expected. If I take this hardware clock offline then the PDCe automatically uses the second hardware clock when it next polls the sources (as configured in specialpollinterval) and this reflects in w32tm /query /source. Again this is as expected. If I then turn off the second hardware clock the PDCe never seems to revert to local /free running clock and will always show the last used hardware clock IP address as its time source even though both are now offline. This is not what I would expect. I would expect when the next poll of the NTP servers is initiated (as per specilapollinterval), that neither would be contactable and the machine would revert to local clock. However if I restart the w32tm service the source is listed as expected (free running clock)

My question is therefore, should this update automatically without needing to restart the w32time service?


Thanks in advance for any responses - I hope there is a definitive on this out there.

Mat

Continue reading...
 
Back
Top Bottom