###BeginCaseStudy###
Case Study: 4
WingTip Toys
General Background
You are a data architect for WingTip Toys. The company uses SQL Server 2012 Enterprise
edition. SQL Server Analysis Services (SSAS) and SQL Server Reporting Services (SSRS)
are installed on separate servers.
Data Warehouse
The company’s data warehouse initially contained less than 100 MB and 100 million rows of
data from only one data source. It now contains more than 10 TB and 10 billion rows of data,
in 25 tables, from 12 data sources.
The largest table in the data warehouse, the factOrders table, contains 5 TB of data. The
factOrders table contains three date keys: OrderDateKey InvoiceDateKey, and ShipDateKey.
The data warehouse server has 1 TB of RAM. Memory usage is currently at 20 percent.
One billion rows of data are added to the data warehouse each month. New data is copied
each night from the data sources into SQL Server staging tables, and existing records are not
updated. The largest data set is order information, which is loaded in parallel into multiple
staging tables, one for each data source. All the staging tables have the same structure and
belong to the same filegroup as the factOrders table.
The dimCustomers table stores customer information that may change over time.
Data Models
You are developing three SSAS databases, as described in the following table.
Reporting
Business users frequently generate reports in Microsoft Excel by using PowerPivot. The
PowerPivot Management Dashboard does not currently display any usage data.
Several SSRS reports exist that use the data warehouse as a source. The data warehouse
queries are aggregate queries that use the factOrders table and one or more dimension tables.
All SSRS data sources use Integrated Windows authentication.
SSRS displays a security access error message when managers run SSRS reports based on the
Operations database.
Reporting performance has become unacceptably slow.
Business Requirements
Improve the query speed of the SSRS reports.
Allow business users to create reports by using PowerPivot and Power View.
Ensure that all users other than business users can view metadata for the Customers
dimension. Ensure that business users cannot view metadata for the Customers dimension.
Technical Requirements
Modify the tables in the data warehouse to minimize aggregate query processing time.
Minimize disk storage in the data warehouse.
Ensure that all multidimensional models process data as quickly as possible.
Create a fact table named factCustomerContact in the data warehouse to store the contact
date, customer key, and communication type for each instance of customer contact.
Store the history of customer information changes in the dimCustomers table.
Move data from the staging tables into the factOrders table as quickly as possible. When
creating dimensions for the date keys in the factOrders table, minimize storage space
requirements and optimize the cube processing time.
Ensure that queries against the Sales database return the most current data in the data
warehouse.
Ensure that the SSAS model of the Finance database does not page to disk or return a
memory error as the size of the database grows.
Create an SSAS monitoring solution that tracks the following data:
• Queries answered per second
• Queries from cache direct per second
• Queries from file per second.
###EndCaseStudy###
You need to implement the date dimension in the Operations database.
What should you do?
A.
Create three database dimensions. Add each database dimension as a cube dimension
by setting the Referenced relationship type.
B.
Create one database dimension. Add three cube dimensions based on the database
dimension. Set the Regular relationship type for each cube dimension.
C.
Create three database dimensions. Add each database dimension as a cube dimension
by setting the Regular relationship type.
D.
Create one database dimension. Add three cube dimensions based on the database
dimension. Set the Referenced relationship type for each cube dimension.