Last week was the MVP Summit. This year included a bunch of very technical discussions about some of the future of the Microsoft Data Platform (a big thank you goes to Slava Oks, Bob Ward, and all the team for a great job). I can’t share much since it was all under NDA. All I can really say about it is that there are some cool things coming.
Fortunately, there are a couple of technologies that are clearly going to have a huge impact on the future of the Data Platform that I can share with you. These technologies are released, and people are already working with them in ways big and small, from development to production. They are notebooks and containers. I’m just barely getting the concepts behind notebooks and the opportunities that they’re going to offer into my tiny brain. So, I’ll just link you to this excellent blog post by Gianluca Sartori on putting Glen Berry’s DMV scripts into a Notebook. It’s a glimpse of what’s possible.
However, I am able to talk about containers. SQL Server has been running on containers for at least 18 months at this point. I’ve been using them myself for just shy of a year. If you have yet to start to explore the concepts around containers and container orchestration systems like Kubernetes, it’s past time for you to get started. This technology is going to disrupt the way we do our jobs in a very big way. I’ve heard them compared to the disruption caused by virtual machines, but containers go way beyond that. What you can do with this technology just changes what we can do, more and faster than just about anything I’ve seen in my 30 years in tech (TCP/IP being released publicly was bigger, but this is big).
What I find most fascinating about containers though is not all the possibilities (and have a chat with the people who’ve been working with them for a while to get some idea of how huge this is, I recommend Anthony Nocentino or Rob Sullivan). No, instead, it’s the resistance. Many people, who have never worked with containers, who haven’t studied them, who have only heard others talking about them, are incredibly quick to say, “It’ll never work.” I’ve heard these voices before. I remember when I first saw a good demo of virtual machines. It was fascinating, except I heard all these people saying “We’ll never have our database in production in a virtual machine. It’ll never scale/perform/availability/something as well as our hardware.” Yet, today, most people are running part, or even all, of their load on virtual machines, in production, and it works fine. The first time I saw Azure SQL Database, I was blown away by the possibilities (and bummed by some of the insane limitations, 2gb, really?). However, behind me, I heard, over and over, “We’ll never have a production system on this because of scale/performance/availability/something.” Yet, today, again, more and more people have at least a hybrid system, split between the cloud and on-premises. I even had one organization tell me that they were getting better performance in Azure SQL Database than they used to get locally (and yeah, maybe there’s a long discussion to be had there).
My point is, before you start saying “never” and complaining about how this technology doesn’t work EXACTLY like the technology you’re used to using, let’s try something different. Instead, it’s probably time to start to look at new technology as what it is, opportunity. There’s a chance that this stuff will fix a problem you have today. There’s a chance this stuff will prevent a problem you’re going to face tomorrow. It’s time to recognize that as a technologist, and if you’re developing, administering, maintaining, architecting databases, that’s what you are, you need to embrace change and get on board with it. Learn this stuff now, while it’s just getting going. That will put you ahead of the curve, but it will also enable you to help your organization to move forward. Trust me, just like virtual machines, just like cloud technologies, containers are coming to your organization. Do you want to be remembered as the speed bump that slowed down successful implementation, or the facilitator that helped to make it happen? Which one do you think pays better?