In terms of technical proficiency, in addition to sharpening their cloud skills, DBAs should also get up to speed on additional technologies that are in the ascent while other recently hot technologies are in decline, said Kline. For NoSQL, Kline said, "Go with MongoDB and/or Microsoft CosmosDB. CosmosDB is compatible with Mongo, while both are powerful database platforms." And, when scaling out across hundreds of clustered computing nodes choose the flexible and capable Apache Spark. Hadoop is "passé," he said.
Kline also advises that DBAs "learn as much about data science algorithms as you can and, even more importantly, how and where to apply them. When coding many data science algorithms, you can select from quite a few programming languages including good old SQL. But the most capable are Python and R. Personally, I recommend Python over R. You can become productive very quickly on Python, but you’ll be wrestling with R syntax for a long time. Yes, R is more sophisticated once you’ve mastered its learning curve, but a lengthy proposition. Furthermore, dip your toe in machine learning principles and applications."
Technical skills are important, but it's likely even more important that DBAs understand the business context in which data is being applied, added McKendrick. “Companies are turning to cloud services and APIs for much of their capabilities—and these services already have technical expertise built into them. At the same time, many cloud services, while far simpler than on-premises systems, are still too complicated for business users to engage with. What is needed is a bridge between the technology and business worlds—individuals, such as DBAs, who understand what cloud services are required and how they fit in to the existing infrastructure, while also providing clarity to their businesses. Further, DBAs need to be able to work with developers and other parts of IT to develop interfaces and dashboards that enables decision-makers get to the information they need as quickly as possible, through one or two clicks on their screens.
‘What is needed is a bridge between the technology and business worlds—individuals, such as DBAs, who understand what cloud services are required and how they fit in to the existing infrastructure, while also providing clarity to their businesses.’ – Joe McKendrick
What DBAs Should Do to Prepare for the Big Data Future
The transformation of the role of DBA is just beginning, experts say, and the time is now for DBAs to step up and get the training and experience that will need to sustain their careers and benefit their organizations. Accepting the growing variety of data and getting better at working with the business are two key aspects to having a long DBA career.
“My advice to DBAs is to embrace the diversity of data stores,” said Harrison. “Be the one who knows how to move data from Oracle to Hadoop, be ready to put your hand up when an upstart database like MongoDB enters the picture. Whatever you do, don’t stay in your comfort zone if that comfort zone is in a contracting market.”
‘Whatever you do, don’t stay in your comfort zone if that comfort zone is in a contracting market.’ – Guy Harrison
On top of that, said Harrison, “If you have an interest or background in statistical programming, now is the time to put it to work by assisting in data science tasks. Languages such as Python and R are particularly widely used in data analytic tasks.”
Since big data tends to be modeled and organized depending on usage patterns, being closer to the business users is critical for data architects to be successful, Caserta said. “We tend to see more denormalization and reduced joins in our data models so getting the data models right is just as important as ever. Going beyond relational, different data storage and database types, including object data stores, column-based, graph, key-value and document all have their own specific data modeling best practices, both logically and physically. The soon-to-be-released ‘Seven Databases in Seven Weeks,’ Second Edition, by Luc Perkins, Jim Wilson, and Eric Redmond, is a book any current DBA should read to gain a practical understanding of NoSQL databases.”
DBAs must understand that organizations need many different types of data and ways to store data in the future, Corey added. “One size does not fit all. They need to gain skills outside of the traditional databases and be open to owning the problem. What is it? How do we use it? Is it accurate? What do we need to retain? How do we keep it accessible? How do we meet business needs?”
DBAs should get to know the people on the business side well, and share their knowledge of data best practices with them, agreed McKendrick. “We're all on this data journey together—business users need to understand data, and DBAs need to understand the business.”
On top of that, advised Reeves, DBAs should automate their current tasks as much as possible. “Whether those tasks are deploying database changes to update a schema or to support a new custom application version or to enable security policy enforcement, at some point, you will be asked to take on new responsibilities. You will then have a choice: work harder or work smarter. DBAs owe it to their friends, family, and themselves to make sure they are able to handle the new challenges ahead in big data without missing piano recitals, Little League games, and date nights.”
This article is part three of a six-part series by editors of IOUG SELECT and Big Data Quarterly on "The Changing Role of the Modern DBA," with three articles appearing on SELECT and three articles appearing on Big Data Quarterly.
The first article on the DBA's role in the new cloud era by Joyce Wells is available on the Big Data Quarterly website.
The second article on on the DBA's role in data security by Michelle Malcher is available at select.ioug.org.