Browse by Tags

Tagged Content List
  • Wiki Page: SQL Server Columnstore Index FAQ

    The SQL Server xVelocity memory-optimized columnstore index stores data by columns instead of by rows, similar to a column-oriented DBMS. The columnstore index speeds up data warehouse query processing in SQL Server 2012 through the current version, in many cases by a factor of 10 to 100. We'll...
  • Wiki Page: SQL Server Columnstore Performance Tuning

    Table of Contents Introduction Fundamentals of Columnstore Index-Based Performance DOs and DON'Ts for using Columnstores Effectively Maximizing Performance and Working Around Columnstore Limitations Ensuring Use of the Fast Batch Mode of Query Execution Physical Database Design, Loading, and Index...
  • Wiki Page: Using a Drop-and-Rebuild Approach with Columnstores for a Nightly Load Scenario

    If you have a nightly load scenario, which is the case for most data warehouses and data marts, then there is no need to worry about the restriction that you can't update a table that has a columnstore index on it. All you need to do is drop the columnstore index, do your updates, then rebuild the...
  • Wiki Page: Parallelism (DOP >= 2) is Required to Get Batch Processing

    Only parallel queries can run in batch mode. Normally this is not a problem because the large majority of servers running SQL Server Enterprise Edition, and being used for reasonably large data warehouses and data marts, have more than one core these days, and queries on larger data volumes will always...
  • Wiki Page: Using Statistics with Columnstore Indexes

    The query optimizer uses statistics to estimate the cardinality (number of rows) in the output from an operator in a query plan. Statistics play an important role in getting a good query plan, so it's worth some effort to ensure that the optimizer has the right information when choosing a query plan...
Page 1 of 1 (5 items)
Can't find it? Write it!