I’m tracking down a web service 400 error, and looking at the HTTP API error logs. I see entries like this:
2011-03-11 19:41:54 10.12.13.131 53292 10.12.12.55 443 HTTP/1.1 >POST /ServiceRequestRest/harpoon/374078EA1FDDDA04AE5D3B79C1662AB1710FAC3F 400 – Hostname –
Nothing too interesting here, except that the time stamp shown 7:41 PM, with today’s date. As I write this post, it is only 1:20pm. Naturally I checked the time the machine provides (Windows 2k3 R2 SP2 box), and it is set correctly.
However, the system event log does contain the following “W32Time” error:
The time provider NtpClient is configured to acquire time from one or more time sources, >however none of the sources are currently accessible. No attempt to contact a source >will be made for 240 minutes. NtpClient has no source of accurate time.
Does this mean that “NtpClient” keeps its own clock, separately from what is available elsewhere in Windows? Why doesn’t it just sync with the time the rest of the OS “knows?” Or does the HttpErr log / NtpClient simply record time in GMT units?
After rebooting this server, the clients were finally able to contact the server (no more HTTP error 400), but the HttpErr log is still producing errors for a time in the future.
On the M$ page that you linked in your question it says: “The Time field follows the W3C format. This field [Time] is based on UTC.”