Q&A: Peeking Below the Surface of the Product Track at MariaDB OpenWorks


Because there’s so much more to a user conference than what you see on the agenda, over the next few days (leading up to the start of MariaDB OpenWorks) we’ll be sitting down with key MariaDB stakeholders to give you fresh insights into the show’s themes and highlights. We’re kicking things off today with Max Mether, MariaDB’s VP of Server Product Management, who organized the topics covered in the Product track at MariaDB OpenWorks.

What will OpenWorks attendees learn from the Product track that they might not discover otherwise?

In the Product track we’re focusing on cool new products and features that are up and coming in the latest releases across our product line. It’s a great way to get an overview across the board (if you go to all sessions) or if you have a specific feature or product that interest you, you can get the nitty-gritty details in that session. It is also a great opportunity to ask questions directly of product management and the developers themselves.

Between the acquisition of Clustrix, the release of MariaDB Platform X3, and the debut of MariaDB Platform Managed Service, you’ve been busy! What are you most excited about at MariaDB right now?

Yeah, we have been busy in so many different areas and we have so many new products upcoming that it’s mind boggling. There are pieces that excite me in each different product (like I’m super excited about Clustrix, about instant ALTER TABLE and the new backup in MariaDB Server) but the most exciting thing to me is that we are starting to have a more cohesive offering and product line focusing on a MariaDB Platform that can then be adapted to different scale, different workloads and use cases. Being able to do analytics and large-scale OLTP from the same platform without having to switch anything for the customer is what excites me the most. If you need OLTP, use MariaDB; if you need OLAP, use MariaDB; if you need large scale use MariaDB. Basically all you need to know as a customer is MariaDB – there’s no need to transfer data back and forth, no need to learn a different language or system…. This is something unique that we are building and very few other products can offer the same. It will take a while to go all the way but we are taking important steps with each release. This is a development I’m very excited about.

In a nutshell, what’s the outlook for machine learning for MariaDB databases?

We are looking at using machine learning in many different ways for MariaDB. A first implementation is for a managed service offering where we can use machine learning to predict a coming performance degradation before it happens. This allows us to recommend expanding a cluster or settings in a cluster ahead of time, basically guaranteeing that the quality of service doesn’t degrade. But there are other use cases under investigation as well.

What question do you hear most when it comes to migrating to MariaDB?

I hear a lot of questions on how easy or difficult it is to migrate. The answer is that it is never easy. We have started a journey where we’ve made MariaDB Server more compatible with SQL standards and other databases out there for quite a while already, starting with features like Window functions, CTEs and adding Sequences, PL/SQL parser etc. However, even if your new database has exactly the same features as your old one there is always a learning curve with a migration. The point is that often the benefits outweigh the costs – you just have to come prepared knowing that there will be some adaption needed. We added all these features to MariaDB not so that we would be 100% compatible but so that it would be easier to migrate and not only to migrate data structures but also to migrate skills. If you are used to using Window functions you should be able to do so in MariaDB as well and not have to think of another way to solve the same problem. That is the main reason we have added compatibility features: so that it would be easier to migrate but also (and more importantly) so that the skills you have with other databases won’t go to waste with MariaDB.

What’s notable about the Product track that isn’t obvious from the agenda?

We have a few overview talks and some deep dives that go into specific areas or features that are new and upcoming or might just be very interesting. We have tried to keep the mixed audience in mind when creating this track, so not all of the sessions are for everyone. However there are several sessions that I am very interested in listening to myself (and I don’t mean my own sessions!) so I think there is something there for everyone. But there are some really interesting sessions in the other tracks too so you are up for some tough choices…. Make sure you read the descriptions beforehand to make sure they are a good fit for you. For example, if you want an overview of the latest cool enhancements in InnoDB you should go to that session, but if you are interested in the InnoDB internals then you might be better of going to the BoF on that topic. (Details on the Birds of a Feather groups will be available at the show.)

We hope to see you in NYC at MariaDB OpenWorks – but if you can’t make it this year, sign up to watch the live-streamed keynote presentation on February 26 at 9 a.m. ET.