9a1292ed   
   From: esquel@sommarskog.se   
      
   Mo (le_mo_mo@yahoo.com) writes:   
   > I have a relatively large table which has around five million records.   
   > When I run a query like:   
   >   
   > Select A, B, C where A='AA' and B='BB'   
   >   
   > the query takes a long time to complete which is slowing down the web   
   > front end.   
   >   
   > My question is how to go about indexing the table to make this query   
   > faster? I have added two (Non Clustered and clustered) indexes to this   
   > table but do not see any improvement in performance. Any   
   > recommendations is greatly appreciated.   
      
   For this particular query, the best index (assuming that the table has   
   more columns that these three) would be one of:   
      
     CREATE INDEX best_ix ON tbl(A, B) INCLUDE (C)   
     CREATE INDEX best_ix ON tbl(B, A) INCLUDE (C)   
      
   But if (A, B) are selective enough, an index on those two alone may be   
   sufficient. I'm a little wary of adding covering indexes for such queries,   
   since if someone later adds the column D to the result set, the index is   
   not covering anymore, just unecessarily large.   
   --   
   Erland Sommarskog, SQL Server MVP, esquel@sommarskog.se   
      
   Links for SQL Server Books Online:   
   SQL 2008: http://msdn.microsoft.com/en-us/sqlserver/cc514207.aspx   
   SQL 2005: http://msdn.microsoft.com/en-us/sqlserver/bb895970.aspx   
   SQL 2000: http://www.microsoft.com/sql/prodinfo/previousversions/books.mspx   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)