|
James K
Flowing Fount of Yak Knowledge
1744 Posts |
 Posted - 02/28/2013 : 07:55:30
|
Sometimes dropping the indexes, inserting the data, and then recreating the indexes can be faster. In some other cases it can be faster to keep the indexes and and insert the data. It depends on how many indexes, how wide those indexes are, how many columns in the index, how the data is distributed etc. So the only way that I know to find out if you would benefit from inserting without dropping the indexes is to do a test.
As for partitioning, take a look at some of the tutorials available online - here is one: http://www.brentozar.com/archive/2013/01/sql-server-table-partitioning-tutorial-videos-and-scripts/ I think Kimberly Tripp has a very good one as well (on sqlskills.com), where she goes into some detail about how you should set it up if one of your objectives is to make OLTP transactions more efficient and faster. |
 |
|
Table partitions in SQL server 2008 Table
