I have "Agent wait time (mins)" in ticket dataset and "Pending status time (mins)" in Update History dataset.
By definition:
Pending status time (min): The time in minutes that tickets were in the Pending status.
Agent wait time (min): The total time in minutes that a ticket was in the pending status. It measures how long agents were waiting for the customer replies.
Which is the best metric to calculate Resolution time ==> SLA.
Should it be Resolution time - Agent wait time or Resolution time - pending status time?
Any clear meaning of why there are two metrics with same definition in two different dataset?
The metrics are essentially the same – they will give you the time spent by tickets in pending status. They are in different datasets, so it would depend on whether you building your query using the Tickets or with the Updates history dataset. If you are looking to exclude the pending time of tickets from the resolution time, then you can create your report under the Tickets dataset since the resolution time metrics are under this dataset; and, in this case, you'll be using Agent wait time in your calculation.