Question

这个查询花了太多时间来执行结果集，有没有办法优化查询

WITH MEMBER PatientName AS [DimPatient].[Full Name].CurrentMember.Member_Caption 
     MEMBER PatientAge AS [DimPatient].[Age].CurrentMember.Member_Caption
     MEMBER PatientGender AS [DimPatient].[Gender].CurrentMember.Member_Caption 
     MEMBER PatientMRN AS [DimPatient].[Medical Record Number].CurrentMember.Member_Caption 
     MEMBER PCPList AS  [DimPCP].[Provider Key].CurrentMember.Member_Caption 
     MEMBER CareTeamList AS [DimCareTeam].[CareTeamName].CurrentMember.Member_Caption 
     MEMBER  ConditionList AS [DimCondition].[Condition Name].CurrentMember.Member_Caption 
SELECT   (([DimPatient].[Patient Key].[Patient Key], 
                   [DimPCP].[Provider Key].[Provider Key],
                   [DimCareTeam].[CareTeamName].[CareTeamName], 
                   [DimCondition].[Condition Name].[Condition Name])) ON 1 ,
{ [Measures].[PatientName] ,
  [Measures].[PatientAge] ,
  [Measures].[PatientGender] ,
  [Measures].[PatientMRN] ,
  [Measures].[Num Of Minutes] ,
  [PCPList] ,[CareTeamList] ,
  [ConditionList] }   ON 0 FROM [NavigateCube] 
  WHERE ( [DimManagedPopulation].[ManagedPopulationKey].&[5059],[DimAnchorDate].[Date Key].&[20160930] )

Answer 1

您可以尝试以下方法吗？无需构建返回维度标题的度量。并且NON EMPTY应该通过仅返回有意义的维度组合来加速它（与我假设的Num of Minutes组合）：

SELECT  NON EMPTY (([DimPatient].[Patient Key].[Patient Key], 
                   [DimPatient].[Full Name].[Full Name],
                   [DimPatient].[Age].[Age],
                   [DimPatient].[Gender].[Gender],
                   [DimPatient].[Medical Record Number].[Medical Record Number],
                   [DimPCP].[Provider Key].[Provider Key],
                   [DimCareTeam].[CareTeamName].[CareTeamName], 
                   [DimCondition].[Condition Name].[Condition Name])) ON 1 ,
{ [Measures].[Num Of Minutes] }   ON 0 
FROM [NavigateCube] 
  WHERE ( [DimManagedPopulation].[ManagedPopulationKey].&[5059],[DimAnchorDate].[Date Key].&[20160930] )

MDX：查询性能

1 个答案: