By Paul Parker, SolarWinds Federal & National Government Chief Technologist
Here is an interesting article from my colleague Joe Kim, in which he explores database heath and performance.
Part of the problem with managing databases is that many people consider database health and performance to be one and the same, but that’s not necessarily the case. Let’s take a closer look at these terms.
Health versus performance: What’s the difference?
Health and performance are certainly closely related, even interconnected. But assuming they are one in the same is potentially a recipe for disaster. If you’re homed in exclusively on your database’s health, you may be overlooking critical metrics that affect your database’s performance. Here’s why:
• Database health is inclusive of data points. When you take into consideration such factors as CPU utilization, I/O statistics, and memory pressure, you can determine if your database is capable of proper performance. But these metrics alone cannot confirm that the system’s performance is running optimally.
• Database performance integrates an element of time measurement to explain how database queries are being executed. It’s this time component that comes into play when talking about true performance.
Diagnosing the root cause: Database performance management best practices
Identifying the true root cause of database performance issues is the goal of every federal database manager. And yet, without the proper metrics in hand, you lack the tools necessary to resolve more comprehensive problems.
That said, let’s take a closer look at some best practices that take into account both health and performance to create efficient, well-optimized database processes.
Acquire data and metrics. You need granular metrics like resource contention and a database’s workload to identify the root cause of a performance issue. Without good, deep intelligence, you lack the ability to troubleshoot accurately and effectively.
Establish meaningful data management. Every database manager has his or her own way of arranging data, but the key is to arrange it in a way that will help you quickly identify and resolve the root cause of a potential problem. Establishing a system that allows you to do so quickly can help keep your databases running efficiently.
Triangulate issues. The ability to triangulate makes it easy to answer all-important questions regarding who, what, when, where, and why. These questions help you determine the details of a performance issue. Understanding who and what was impacted by poor performance and what caused the impact are important to know.
Review execution plans. Query optimizers are critical database components that analyze Structured Query Language (SQL) queries and determine efficient execution for those queries. The problem is that optimizers can be a bit of a black box; it’s often difficult to see what’s going on inside of them.
Establish a baseline. It’s impossible to tell if your database isn’t performing optimally if you lack a baseline of normal, day-to-day performance.
It’s a marathon, not a sprint
IT pros want their databases to be in good health and to perform optimally. While both are equally important, it’s the end result that matters.
So, make sure you are looking at all criteria of database health and performance. If you are deploying the best practices and tools to help ensure the overall health and performance of your database, your stakeholders will thank you for it.
Find the full article on our partner DLT’s blog Technically Speaking.