Over a year ago, we applied a fix related to this on our domain controllers because of intermittently reported client problems. This fix was the most significant finding out of Microsoft AD RAP as a service engagement that we had.
Turns out that down in the deep details of the time settings there’s a lot that can cause problems. And Microsoft deployed poor values for many of those settings for prior OSes, some of which got encoded in the default group policy values. We turned on those group policy settings back when the values were really bad, and of course, those bad values were still in place because nothing will automatically change an existing GPO setting.
Here’s a table summarizing the craziness:
|2008R2 Admx Defaults||2008 / 2008R2 DC Default*||2008 / 2008R2 Member Defaults||2003 Adm Defaults||2003 DC Defaults|
* These are the settings you want on you DCs as they are the OS defaults for a DC
On your DCs, you want to apply what’s in the column labeled “2008/2008R2 DC Default*”. To correct our problem, I turned off all the GPO time settings. Then I turned them back on. Then I fixed the only setting whose (GPO) default value doesn’t match the OS default value for a DC (PhaseCorrectRate).
This fixed all of the time events being raised in our domain.
YMMV, but I expect this will likely help quite a few folks.