The DBA’s Dilemma: How to Upskill and Stay Ahead

The joke goes that no one knows what DBAs do. That’s because the position is rapidly evolving, and the job description varies by the shop. At some companies, DBAs write all the stored procedures. Elsewhere, they don’t perform such duties—developers do.

Most DBAs administer databases within hybrid IT stacks. Some DBAs are still at the data center with the server racks some 20 feet away. Others work only with cloud-based databases. All DBAs have one thing in common, though: When something goes haywire with a database or a SQL query, everyone expects them to fix it.

However, DBAs aren’t magicians. In my experience, there have been unrealistic expectations of what they can do. When I was a DBA, management expected Superman-level results; we needed to solve every problem. Sure, I fixed countless issues, but sometimes I also had to level-set expectations.

The Upskill Challenge

As cloud technologies, and machine learning and AI revolutionize the industry, DBAs must upskill. Of course, some factors stand in the way. Usually, there’s a great deal more legacy code being maintained than there are people building shiny new things. There is also the usual lack of resources. In addition, training budgets come and go, and, of course, lack of time is the standby classic—and that’s why DBAs attend user groups at nights and on weekends.

But, if DBAs can’t or won’t upskill, they’re stuck and can expect to continue in their roles until those roles no longer exist. This is why they should constantly be thinking about what to do next—their next pivot. The pivots can be slight, but they need to be in the direction of opportunity.

Fall In Love With Data Engineering

Traditional DBAs who have spent a lot of time with all the nerd knobs at their disposal need to think more about the movement of data. Instead of thinking in terms of data administration—and running the risk of becoming redundant—they should become more familiar with how data moves in and out of their enterprises.

For example, if your company opts for a technology such as Amazon RDS—which can perform the backups DBAs do, along with auto-tuning and auto-scaling—there’s a risk of getting outmoded by tech. When DBAs perform their operations, it can take time to configure and get set up. But with RDS, it’s just a matter of flipping the switch.

That’s why DBAs should instead focus on the data aspect itself. There will always be a problem with the report. Focus on the data aspects, and you can be the one looking at the report and seeing the zeros showing up as nulls, making the sum wrong. With your expertise, you can be the one to figure out how to make it right. Traditionally, that might not be what DBAs do—unless they’re the ones writing procedures. But what are you doing talking about tradition, anyway?

To my way of thinking, expertise in data engineering makes you a highly desirable employee. I say this because an incredible amount of data out there hasn’t gone to the cloud yet. It must get there—and there’s nothing wrong with you becoming the person who is the expert on how to do it. There’s so much within data engineering: You can learn Python and learn how to work with data, not just SQL.

Your days of tuning queries are ending one page at a time. Microsoft is actively building self-tuning AI into SQL Server. It will recognize queries and then run as efficiently as it did last time. It’ll also build you a different execution plan—and it’ll run better than the last one. No DBA will be involved in that effort.

And, at the end of the day, it will be the DBA who is the scapegoat for any loss of data. It won’t be your fault that the data is gone, but it will be your responsibility. That’s why you must understand the cloud heavyweights, Azure and AWS. You need to understand how the cloud may or may not be doing your backups.

It helps to know about application performance monitoring, or APM, too. Being able to tie a price to a database or query helps management understand the actual cost. APM is vital when you get into cloud-native apps because if you have a query running a million times, it is consuming resources that could be costing you $1,000 a month due to a simple coding issue or a change in architecture.

Understanding cost and cost optimization is part of being a data professional. DBAs will keep their titles, but they should consider themselves data professionals—with a focus on data engineering.