We're working with a new install of 9.1.2 and during some initial testing of a few parallel jobs we are noticing that the elapsed time is incorrect on all jobs. In the below example the job started yesterday afternoon and only ran for a few minutes. The finished time was a value that didn't yet exist and the elapsed time showed 5 hours 4 minutes 22 seconds. The actual run time was only a couple of minutes. Has anyone experienced this before or can anyone point us to a location that we need to look further into? Thank you!
The behavior I have seen is that the log entry timestamp is NOT the time in which the event happened. It is the time in which the event was LOGGED. (there is a difference)
In your case... you should also look to see if you are running in a cluster and ensure that all of your servers are using the same timezone setting.
several hours worth of difference... (look for TZ setting)