Issue 8  May 4 2021

Think back to a time when you automated something. It doesn't matter what, or how. Pick an example of a task you did manually and invented a way for the task to be automated. And yes, feel free to consider "buying a piece of software" as an example of automating a task.

If you were the person responsible for performing the manual task, you have the necessary knowledge on what to do should the automated process fail. Hopefully you aren't the only one with this knowledge. Perhaps you have it documented for others. This way, someone else can step up and perform the manual tasks.

I mean, it's the same idea with Bob Ross and painting. He documents exactly what needs to be done and makes it easy for any one of us to pick up a brush and create a replica of his painting. Right?

No, of course not. Bob has knowledge and experience, which is crucial to perform the task. Unless others have experience performing your task manually, the documentation is likely to fail at some point.

I had a similar situation years ago. A disaster struck and I needed to restore the master database. Unfortunately, we used a 3rd party vendor for our database backups, and their automated process of restoring master failed when I needed it most. So, I performed the tasks manually, and documented the process for my team. I then asked them to practice restoring master so they would be familiar with the manual process, should it be necessary again in the future.

With advances in data and database technologies, I see automation removing the need for many manual tasks. I think this is a great thing. If I never have to rebuild an index again, I'm good.

However, the next generation of DBAs won't even be called DBAs. They will be a hybrid of administrator, developer, analyst, and scientist. That's the real future of databases 10 years from now. As the tech changes, so will the roles and responsibilities. Automation frees up our time to do other tasks. Instead of needing four experts in each area, you only need one person that has cursory knowledge of four things.

The trouble, as I see it, comes into play when the tools break, and that person needs deeper knowledge and experience.

When the machine stops, you need to know how to repair it. Otherwise you'll look pretty stupid staring at the machine not working, and not knowing how to make it work.