Because this event’s value for monSysWaits.WaitTime is measured in seconds, the value for WaitTime for this event should be much less than the value for Waits (an average physical read should be 2–6 milliseconds; more than 10 milliseconds is considered slow). A high average physical read value may indicate poor disk throughput performance. Query monIOQueue and monDeviceIO to identify slow or overloaded disks.
A high value for Waits, regardless of the value for WaitTime, may indicate that query plans are not as effective as they could be. If you encounter a high value for Waits, a table scan or Cartesian product may have occurred, or the optimizer may have selected a bad plan, due to bad, stale, or missing statistics. Consider adding an index on specific columns to the table on which this occurred.
A high value for Waits can also indicate the data caches are too small, with active pages first pushed out and then reread. Query monOpenObjectActivity, monProcessActivity, monDataCache, monCachPool, and monProcessObject to determine how to proceed.
