I have a search like this:
index=\* earliest=+1m latest=+30h sourcetype="WinEventLog:Sys*" Message=\*Upgrade\*
It looks at events in the future.
I have a customer that has some events in the future and it looks like it may be related to a restart of the system, maybe only after upgrading the OS.
And it may be related to the machine not having the correct time set when it starts logging, and only after getting the NTP server it logs correctly.
Any of you guys have had same issues? Any comments / feedback would be appreciated.
My guess is it is a Windows problem, not a Splunk problem.
If you are using Splunk's built-in
WinEventLog input, the only way for this to happen is for the system clock on the forwarder to be incorrect because this is the time that is used for all events and the Indexer does not have anything to do with modifying it.
Yes, so if the server after a restart, maybe only in context of an OS upgrade, has a wrong system clock set until it gets a hold of its NTP server, it could yield some events that are off.
I may have been unclear in my question above. The "looks like it may be related to a restart of the system". The "system" is the OS the forwarder runs on.