Building Scalable Analysis Services Applications - Microsoft SQL Server 2008 Analysis Services - page 448

The advantages to this solution are as follows:

. Simplified multidimensional model —With a single computer, you don't need to

distribute and integrate data among different computers.

. Simplified management model —You manage only a single Analysis Server. All

data-manipulating and data-loading scripts are targeted at a single computer. Moving
data from a staging server to a production server is a simple process. With a scale-up
approach, it's relatively easy to back up your database. It's easier to restore the appli-
cation state from the backup. It's also easier to monitor your application. With a
single computer, it's easier to find causes for slowdowns.

. Simpler network configurations —You don't need to configure the network to sup-

port a distributed infrastructure.

But this solution has its disadvantages:

. Hardware cost —At a certain point, scaling up to move your application to more

and more powerful computers gets you into the price range of enterprise-level hard-
ware. Enterprise-level hardware that enables you to scale to large data volumes is
substantially more expensive than lightweight database hardware.

. Single point of failure —A single computer with attached storage is prone to failure

more than the scale-out solution, which runs on multiple computers. To ensure high
availability for your system, you need to pay an additional premium for fault-toler-
ant or highly reliable hardware.

. Application growth —It's hard to predict the growth of data volumes for your appli-

cation or the number of users your application will have to support. With a scale-up
approach, you have the burden of having to choose the right hardware for your
application. Capacity planning is a delicate and risky task.

. Scalability limits —Relying on a single computer means that at some point you

might run into the limitations of a single hardware component. For example, you
might run into bandwidth limitation with your hardware storage solution.

. Application configuration cost —Running Analysis Services on enterprise-level

hardware requires additional application configuration tuning.

The Scale-Out Approach

With a scale-out approach, your application and your data are distributed across several
computers that work together to load data into the system and answer user queries. To
scale your system, you add additional computers to your application. The disadvantages of
a scale-out solution are the opposite of the advantages of those of the scale-up solution. In
the following list, we discuss only the advantages:

. Hardware costs —To scale out your application, you use commodity hardware that is

a fraction of the cost of enterprise-level hardware.