carloscastilla - Fotolia

MariaDB SkySQL enables cloud-native database as a service

MariaDB brings back an old name for its cloud-native database-as-a-service platform. MariaDB CEO Michael Howard details how he's looking to compete against the big cloud providers.

Back in 2013, database services company SkySQL merged with the key backers behind the open source MariaDB database project to create a new combined entity known as MariaDB Corp.

Since then, the SkySQL name itself has been dormant, with MariaDB being the core company and brand name, but that changed on March 31 with the general availability launch of the MariaDB SkySQL cloud-native database as a service (DBaaS). The MariaDB SkySQL service extends the company's X4 database that was released on Jan. 16.

In some ways, the MariaDB SkySQL service is the fulfillment of an original vision that is now finally being realized by MariaDB CEO Michael Howard. That vision is making it easier for users to deploy and consume MariaDB.

In this Q&A, Howard discusses the need for cloud-native DBaaS and where the market is headed in the turbulent year ahead as the world grapples with the COVID-19 pandemic.

What impact has COVID-19 had on MariaDB as a company? And do you expect that current work from home climate will drive demand for SkySQL?

Michael Howard, CEO, MariaDBMichael Howard

Michael Howard: Pervasive problems require pervasive solutions. My hope is that SkySQL will empower and trigger creativity for everyone -- working from home, working to keep the lights on, working to help conquer COVID-19. This is why we wanted to get SkySQL out now, to give that edge to spirit and ingenuity to help make life a tad simpler in tackling serious problems. We are here to help.

Is MariaDB SkySQL just the X4 release, in the cloud? Or is it something else?

Howard: It's more than just X4, it's a very interesting amalgamation of power. Yes, it does have X4, whose primary distinction in the marketplace is to have analytic and transactional processing all in the same database. X4 is also more than that, with the way in which data is propagated from the row store to the columnar store in an approach we call smart transactions.

Typically what has to happen is that data has to be propagated to a separate data warehousing instance, like a Redshift or Snowflake. Here it's part and parcel of the same database. So that's unique, and that's manifested in SkySQL. That's no easy task to do and we automate things like block storage as well.

On top of that is Kubernetes. We spent a lot of time optimizing the Kubernetes footprint to really handle a persistent technology like a database.

Why did you decide to bring back the SkySQL name for the first time since 2013?

Pervasive problems require pervasive solutions. My hope is that SkySQL will empower and trigger creativity for everyone -- working from home, working to keep the lights on, working to help conquer COVID-19.
Michael HowardCEO, MariaDB

Howard: If you go back to 2013, all the different parts were consolidated under one roof. At that point SkySQL was one of the components whose aim -- though they never got it done -- was to support cloud uses of MariaDB and MySQL.

That was the innate goal or definition of the SkySQL components that were all consolidated in late 2013. The combined company was just called MariaDB, because SkySQL just didn't seem like a good name for a database company whose database is MariaDB.

About four and a half years ago, when we started on this journey to build a MariaDB cloud database, I thought SkySQL was an excellent name for our cloud, so that's why I resurrected it.

Why did it take so long to build a cloud-native DBaaS? What were the key challenges?

Howard: We tried many different techniques. We started off with a physical instantiation. But it was about two years ago when we saw the opportunity with Kubernetes.

Because of MariaDB's community and global footprint it was clear to me that we would have to be multi-cloud. Doing that physically and having a portal on each of the clouds was just an unnecessary evil. So after starting with using virtual machines and physical instantiation methods, we moved to the Kubernetes method.

What are your concerns and views on the risk of big cloud providers cannibalizing other databases?

Howard: My initial ideology on the topic is that the major cloud vendors are 'strip mining' and not ever giving back to the community. But now I think the opposite is going to happen.

People that are using these strip-mining versions of MariaDB and MySQL will come to SkySQL because they will recognize not only the authentic technology, but one that's just flat out better.

So, I give my hat's off to Jeff Bezos and company [at AWS] for really starting to get the ball rolling in terms of making the public cloud a staple item in our IT environment. But what they don't know how to do is focus and not only focus, but they are not able to have the quality around it. So, they started an addressable market. That's wonderful for us, because we already have companies large and small, already moving to SkySQL.

Dig Deeper on Database management

Business Analytics
Content Management