Hi All,
I am on SQL Server 2008.
I have a table with 300M rows, with one clustered and one non-clustered index.
After collecting stats on the non-clustered index (sample 25 percent), I did dbcc show_statistics. First section of the stats looks like this
All density Average Length Columns -------------- -------------- ---------------------------------------- 1.8214936E-3 5 column1 2.0837042E-8 20.037888 column1, column2 5.2660818E-9 28.037888 column1, column2, column3 5.2656612E-9 36.037888 column1, column2, column3, column4 4.6398854E-9 41.037888 column1, column2, column3, column4 4.6398854E-9 45.037888 column1, column2, column3, column4
What exactly does "All density" stand for? Is it "for one value of column1, how many records are you likely to find?"? May be not because in this table, for many values of column1 we have hundreds of thousands (even millions of) records. Here the density is 0.00182149936 (for only column1), what does that number stands for?
Thanks in advance