BACKGROUND IMAGE: Vertigo3d/iStock
Database bottlenecks put applications at risk if left untreated
Even in the big data era, with Hadoop clusters and Spark systems on the rise, databases remain the lifeblood of IT architectures in most organizations. As a result, database bottlenecks that cut off the timely supply of data to end users can cause the applications at the heart of business operations to stop beating normally. And that condition isn't conducive to long-term corporate health.
Database administrators (DBAs) and software programmers can take steps to avoid database performance issues. For example, in a series of three blog posts published in August and September of 2016, consultant Craig S. Mullins detailed 12 rules of thumb for writing good SQL code to help produce database applications that run efficiently. But, in a follow-up post, Mullins cautioned that those guidelines aren't a guaranteed cure for what can ail databases.
"Rules of thumb are great, but let's face it, we don't always follow them. And then we get performance degradation," he wrote. When that happens, he added, DBAs are saddled with the tough task of having to identify the root causes of performance problems and prescribe a treatment plan to resolve them.
That work can be done manually -- or with the aid of database performance management tools designed to automate monitoring of resource utilization and diagnosis of database bottlenecks. Such tools point to "locking, I/O, CPU and other problems as your applications are running," potentially streamlining the work required to get data flowing properly, Mullins wrote.
But there are different types of software that focus on separate aspects of the process. To help make sure you select the right tool, or tools, to meet your organization's particular database-doctoring needs, this buyer's handbook offers advice from Mullins on the key features and functions to look for across the different product categories.