sum (wait time over a period) / elapsed time of that period = 'Average Users Waiting Counts' . This metric represents the average number of users that have made a call to the database and that are waiting for an event, such as an I/O or a lock request, to complete. If the number of users waiting on events increases, it indicates that either more users are running, increasing workload, or that waits are taking longer. If that waits are taking longer then you should worry about it. Try to use the ADDM or ASH reports for a more detailed analysis.