3 Choices You Need to Stop Making with SQL Servers

 Choices You Need to Stop Making with SQL ServersImagine your office without a database management system. It’s the stuff of nightmares, isn’t it? You’ll envision a slow, chaotic organisation that is overrun with projects it cannot handle and, more than likely, security issues it cannot manage. Fortunately, you can wake up from the ‘nightmare’ and walk into a productive office today — or will you?

It Still Needs Optimisation

SQL Servers, much like other database management systems, still require fine tuning. Performance tuning for SQL Server ensures optimisation no matter what occurs in organisations; critical changes, such as an increase in data, usage patterns and upgrades in the system will have an effect on server performance. And server performance has an effect on employee productivity, and the business.

With adequate performance tuning, your business achieves two objectives:

  • Enhance user experience, boosting productivity, and
  • Prevent future issues that delay tasks and projects.

So how do you fine tune SQL Server performance? Avoid these three decisions that will make your database management system ineffective:

1. Combining data and log files into one physical drive.

Combining your data and transactional log files in one physical drive isn’t a hard and fast rule. But separating them will optimise server performance, and enable recovery in the event of a disaster. If your employees cannot access the drive holding your database files, they can still access the transaction log, if the two are on different drives.

2. Neglecting I/O bottlenecks.

Worse yet, you’re not aware your system is experiencing I/O (or input/output) bottlenecks. You don’t know the queries that are leading to the slow performance. You can use DMFs (dynamic management functions) or DMVs (dynamic management views) to access statistics about your system. In doing so, you determine its condition and decide how to proceed from that point: add new hardware or simply locate bottlenecks, and resolve accordingly.

3. Failing to look at your ‘wait statistics’.

In business, waiting can be a good thing or a bad thing. When technology’s involved, it’s generally the latter part.

SQL Servers will track incidents that hold up requests; the system records the length of the wait, and the reason behind the wait (also called the wait type). These are your wait statistics.

So you’ll have data to tell you what is actually driving down your server’s performance. Why are there requests being held up? What are the wait types? By looking at your wait statistics, you’ll have a better idea of how to fine-tune SQL Server performance.

Technology should work for your organisation. But it still requires adequate maintenance. When it comes to your SQL Server, fine-tune the system to ensure consistent performance.