HiI'm not that experienced in MDX, and having some trouble with a query I'm writing.Here it is:[code]WITHMEMBER Measures.GIDCountASCOUNT( NONEMPTY( ([Dim Employees].[GlobalID].Members- [Dim Employees].[GlobalID].[All]), ([Measures].[Outliers - ACD Calls], [Dim Phone Logins].CurrentMember) ) )SELECT {Measures.GIDCount,[Measures].[Outliers - ACD Calls]} ON 0,NONEMPTY ( ( [Dim Phone Logins].[Phone Logins].Members - [Dim Phone Logins].[Phone Logins].[All] --[Dim Phone Logins].[Phone Logins].[Phone Login].&[874007] ) * ([Dim Employees].[GlobalID].Members - [Dim Employees].[GlobalID].[All]) , [Measures].[Outliers - ACD Calls] )ON 1FROM [TheCube]WHERE([Dim Call Distributors].[Call Distributors].[Call Distributor].&[88],[Dim Call Routes].[Call Routes].&[335361],[Dim Local Time].[Local Time].[Days].&[20150501])[/code]You can see that on the Rows (1) axis, I'm cross-joining the members of two dimensions (PhoneLogin and GlobalID), and getting rid of the combinations that don't have any value for the measure concerned.On the 0 axis, I show the measure value.The purpose of the query is to work out where one member from PhoneLogin has data (within the slice) for more than one GlobalID. This is a data problem which I'm trying to diagnose and debug.So I have this calculated member that counts the non-empty (on the basis of the measure) GlobalIDs for the current PhoneLogin (again, within the slice).Here's the problem.a) When I test this with one member from PHoneLogins (you can see this commented out), it comes back blindingly fast with the correct values.b) When I test it with the entire PhoneLogins dimension, but without the calculated member, again it works blindingly fast: out of c. 600,000 members in dimension PHoneLogin, and c. 300,000 members in dimension GlobalID, it returns just the 400 or so rows that matter.c) When I test it with the entire Phone Logins dimension, and with the calculated member, it just runs and never finishes.What am I doing wrong?thanks for any ideas!
↧