Evolution
I as a DBA should actually say that is not wise, that you always need a DBA to keep things running, fix performance issues, install patches, etc. From the actually in the previous sentence, you might have already inferred that I look at this development very positively.
People work is always a source of errors and disruptions in IT. Even seasoned technical application administrators, system administrators and even DBAs sometimes make mistakes. Sometimes you don’t notice, but sometimes it can have disastrous consequences. What if we could minimize these kinds of mistakes by having a database, in this case, run completely autonomously.
The database takes care of performance tuning, the database takes care of installing security patches, the database even takes care of giving you extra CPU, memory, disk and network capacity when needed. If needed, it could even use less capacity because you completely fail to use the resoures you think you need and thus pay way too much.
Am I not putting my own job on the line with this kind of statement? No, definitely not. As a DBA, you will be using the knowledge you have of a database platform in a different way than the “standard” maintenance stuff that runs autonomously in this case.
You are less concerned with infrastructure issues (CPU, memory, disk space), but you are more concerned with database design.
You’re less concerned with rolling out patches, but you’re more concerned with analyzing the data.
You’re less concerned with ensuring availability, but you’re more concerned with how to keep the database available.
You’re less concerned with performance tuning, but you’re more concerned with securing the database.
You’re less concerned with putting out fires, but you’re more concerned with supporting developers to prevent fires from starting.
I can’t wait for when Oracle 18c comes out to see what conveniences it will bring us as a DBA. And I’m incredibly curious to see how current “production” DBAs will pick up on this shift in their world.
In any case, I am aware that the world around us is changing and that we need to change with it, because there is a very good chance that actually sitting at the controls of databases could be over in 10 years or so. Dba’s will have to find their added value and convince the companies they work for that they are still of great value.
Last Friday at our office in Naarden we had a session on multi platform database management, there the subject of automation also came up. There too we came to the conclusion that the world of database management is changing enormously.
A DBA will no longer do his work tucked away in a corner of the building. He/she will have to interact much more with different parties, be familiar with the world around him (OS, network, cloud, storage), be able to come up with innovative solutions and provide companies with solid advice.
A new definition of DBA could be: “A communicative spider in the web between application, hardware, network, security, BI and the business with knowledge of multiple database platforms, both cloud and open source.” Larry Ellison said the following about this: “You’ll see a migration, an evolution of database skills, where you’re focused more on database design, schema design, different kinds of data analytics including machine learning, setting the policies as to what is mission critical, what requires disaster recovery, figuring out those policies.”
I think that in the future there will be more and more automation and with the advent of “autonomous” systems there will be less and less repetitive tasks performed by DBAs and the DBA will become much more of a controlling factor of these types of systems.