識別前 20 個在讀取 I/O 時耗用最多資源的查詢

識別前 20 個在讀取 I/O 時耗用最多資源的查詢

SELECT TOP 20 SUBSTRING(qt.text, (qs.statement_start_offset/2)+1, ((CASE qs.statement_end_offset WHEN -1 THEN DATALENGTH(qt.text) ELSE qs.statement_end_offset END - qs.statement_start_offset)/2)+1), qs.execution_count, qs.total_logical_reads, qs.last_logical_reads,qs.min_logical_reads, qs.max_logical_reads,qs.total_elapsed_time, qs.last_elapsed_time,qs.min_elapsed_time, qs.max_elapsed_time,qs.last_execution_time,qp.query_planFROM sys.dm_exec_query_stats qsCROSS APPLY sys.dm_exec_sql_text(qs.sql_handle) qtCROSS APPLY sys.dm_exec_query_plan(qs.plan_handle) qpWHERE qt.encrypted=0ORDER BY qs.total_logical_reads DESC

此結果依照成本來排序 產生的結果 xml，可儲存成 .sqlplan，再以 management studio 打開，可檢視其 execution plan. 取自 ：http://www.microsoft.com/technet/technetmag/issues/2007/11/SQLQuery/default.aspx?loc=zx 之 http://www.microsoft.com/technet/technetmag/issues/2007/11/SQLQuery/default.aspx?pf=true&fig=true&loc=zx