“Just use partitioning” is the answer you hear, when you need to manage very large data sets in your Data Warehouse. But how do you design and implement it? We will walk through different ways to design partitioning, including layered partitioning.
This session will discuss what a modern strategy for data warehousing can be in this era, considering how the use of technologies like PowerPivot or Analysis Services Tabular affect the way you should model your data.
In this session, we are going to explain and test different DW features in SQL Server 2012, including star join optimization through bitmap filters, table partitioning, window functions, columnstore indices and more.
: We are looking at the dawn of a new workload: BigData. In this session I will talk about what BigData is, and which BigData technologies Microsoft are working on.
Users love flexible analytics but hate to wait for the data to be loaded into a traditional data warehouse. John will describe how to build an infrastructure to support real-time loading of your OLAP cubes so your user's get exactly what they want
SQL Server FastTrack References Architecture gives you the best practices and reference architectures from different hardware vendors to help you create a balanced environment for your data warehouse workload.
This session looks at some of the different methods available to load slowly changing dimension data into a data warehouse, and compares the relative performance given different data scenarios and traditional storage compared with FusionIO
This talk will describe how the new ColumnStore index technology in SQL Server 2012 makes queries go faster. Covering details of the storage and execution model, how this model interacts with modern CPUs to deliver significant performance benefits.
In this first of two sessions, we review the architecture of SQL Server and its BI components and deployment options for optimal performance. We'll also discuss how to optimize data warehouse load operations.
Snapshots without snapshots...is that possible? Take a "Classic" snapshot fact table, add some temporal data theory and you'll get a new fact table than can store snapshot data without doing snapshots. A life saver when you have a lot of data.