Why would SQL Anywhere Monitor 126.96.36.19910 widgets show vastly different metric values in a Graph display versus a Table display?
The following screenshot shows "CPU Usage 0% at 10:03" in the "CPU Usage History" graph on the left, as opposed to 38% in the "Key Performance Metrics" table on the right.
(more information after screenshot...)
In actual fact (as shown by the following Foxhound screenshot) the CPU Usage has been over 20% on a continuous basis for a long time. This isn't happenstance, the application is pounding the database without pause... there is no way that a CPU usage measurement of 0% could be taken during that period.