SQL Server 2012: performance column index versus B-tree

One of the advantages of the columnstore index is that the data of one column is stored โ€œnext to each other on diskโ€. This means better compression and faster read times.

Nonetheless; when using a B-tree (regular non-clustered index) are the leaves of this tree not the data itself? Therefore, when I make an index in column A, will all the fields of column A be displayed on the disk parallel to the table? This means that data can be compressed and can be read quickly.

In short: when exactly is columnstore-index more efficient than a B-tree in one column?

Thanks in advance!

+4
source share
1 answer

So when I do an index in column A, aren't all the fields in column A displayed on disk in parallel with the table?

Yes, column A will be sorted. But all other columns will not. The column store index stores all columns in sorted order (in pieces of 2 ^ 20 rows).

+1
source

Source: https://habr.com/ru/post/1403926/


All Articles