Administering Core Business Intelligence Services - Microsoft SQL Server 2008 R2 Administration - page 141

For a secure deployment:
Grant SharePoint permissions to individuals and groups; this is necessary to
enable access to sites and content.
Change the PowerPivot service application's pool identity to run under a
different account.
F On SharePoint Service Applications page in Excel Services Settings create
additional trusted sites in Excel Services, so that you can vary the permissions and
configuration settings that work best for PowerPivot data access.
Implementing Scale-Out of SQL Server
Analysis Services
Scalability describes a system's ability to utilize additional resources for an increase in
system capacity to perform additional work. The system capacity is important in achieving
maximum scalability. There are two types of Scalability: 'scale-up' and 'scale-out'.
F Scale-up: It enables the data platform to support the increasing numbers of users,
or applications, taking SSAS abundance of internal optimization and performance
tuning techniques to support the system resource utilization.
F Scale-out: It provides the path to inherent limitation in a scale-up architecture,
increasing server performance by fine-tuning the worker threads, pre-allocation of
memory, and disk resource allocation. It purely depends on the technology limits of
Symmetric Multi-Processing (SMP) scalability.
While most server applications today can scale up to eight logical processors, it is possible to
ramp up Analysis Services with 16, or 32 logical processors. This can be done by adjusting the
size of the thread pools for the formula engine and the storage engine. It is vital to optimize
the system configuration for the best possible performance, which includes the SAN storage
optimization to realize high scalability with SSAS. Once the scale-up is completed up to a
satisfactory level, then scale-out can multiply the scale-up gains.
In this recipe, we will look at how to implement the scale out of SQL Server Analysis Services.
Getting ready
The server configuration is an important aspect to design a scale-out implementation and the
following list highlights the information that is essential to implement a scale-out process:
F Core SQL Server database engine server x 2
F Analysis Services processing server x 1
F Analysis Services query servers x 3