Step 2: Understand your workload

Before you make a hardware decision, you must fully understand your workload and capacity needs. Contributor Hilary Cotter offers his advice.

One of the most important factors in making hardware decisions is to have a good understanding of the workload. Will it be write intensive or read intensive? Will there be significant processing or be less processor intensive? Will these characteristics change over time? How large will your database grow every year? Will any other applications be running on SQL Server that will affect the performance?

You must be aware of these factors before you can deliver a hardware solution that offers optimal performance for the lifetime of your application -- typically it's two to five years before the application has to be rewritten to take advantage of the latest programming and hardware technologies. With this relatively short lifecycle, also keep in mind that the machine may need to be repurposed for other applications after it can no longer meet the needs of the new-generation applications. In general, when you spec out hardware, it is best to be generous in your estimates rather than deliver hardware recommendations that meet the current estimated requirements.

Without a representative load, it is hard to quantify the performance characteristics of the application and then adequately size your hardware. You must use the best guesses developers provide at design time. Most hardware vendors offer white papers that show performance benchmarks for their products under different types of load. But experience is still the best tool you have for making hardware selections. It is wise to select hardware that is componentized and has room to grow ('significant head room'). Make sure the hardware you select has RAM and processor-expansion capabilities.

Spec your SQL Server hardware needs


Hilary Cotter
Hilary Cotter has been involved in IT for more than 20 years as a Web and database consultant. Microsoft first awarded Cotter the Microsoft SQL Server MVP award in 2001. Cotter received his bachelor of applied science degree in mechanical engineering from the University of Toronto and studied economics at the University of Calgary and computer science at UC Berkeley. He is the author of a book on SQL Server transactional replication and is currently working on books on merge replication and Microsoft search technologies.

Dig Deeper on SQL Server Database Modeling and Design

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.