Improving Database Change: Q&A with Datical's Dion Cornett


Database release automation provider Datical recently announced the appointment of Dion Cornett as president. Datical is built on top of Liquibase and is the primary maintainer of the open source project that enables application teams to version, track, and deploy database schema changes. Cornett, who joined Datical with more than a decade of open source leadership experience from positions at Red Hat (which was acquired by IBM) and MariaDB, talked with BDQ about why it is critical  to include the database as a key player in DevOps processes.

The demand for speed in application development and deployment is increasing. How is that impacting database code change and deployment processes?

Cornett: Every meaningful application is using data to some extent. If you are developing applications, you are finding that database code can be a bottleneck. We see this everywhere—in hundreds of customer and prospect conversations. According to a 2019 Datical survey, 92% of app developers reported difficulty in accelerating the release of applications because of the bottlenecks they run into with database schema changes. If you are not keeping up on the database side, then you are not getting the full benefit of speedier app development.

Why is it important to introduce automation into these processes?

Cornett: You have to be particularly careful with the database. If you are building an application and there is some issue with it, you can blow up the application and build a new one. That is not a luxury you have with databases. Databases are stateful. Even something simple such as adding a column can be problematic, and blowing up the database and trying to start over can mean very significant cost to the enterprise. That is why it is more important than ever to carefully automate, reduce friction, and mitigate risk around database change deployment.

How is automation being accepted in the database management area?

Cornett: People have been so successful with DevOps, which was created to develop applications faster. But someone forgot to tell the database, and so I think we are sort of at the fateful point in the evolution of IT at which people are acknowledging that this is a bottleneck and looking for solutions to make database deployments more nimble, scalable, and reliable.

What are some of the key areas you will focus on at Datical?

Cornett: First and foremost, I want to maintain the incredible culture that the company has built here. When I was at Red Hat I was involved in evaluating dozens of companies and had a chance to see the culture across a wide breadth of startups. Datical has a great culture here. People work together well and it is a meritocracy in terms of trying to advance the mission, and then you couple that with tremendous empathy for the customer and an emphasis on creating customer value. And so maintaining those values is important.

What do you want to change?

Cornett: Where I want to make an impact and drive some change is in making available a greater array of innovation to solve the challenges with the database and the real mechanism to do this is leveraging the superior production model that represents open source—being able to engage with and be a good steward of the open source community, and it is a very robust one around Liquibase already. There have been 13 million downloads over the last 12 months, there are hundreds of contributors, and thousands and thousands of commits.

By paying attention to that community and ensuring that feedback is quickly incorporated into what we are trying to build, we are going to create better software. People that want database change automation will future-proof their software by relying on Liquibase. That is really where my expertise is coming in—working with that community to create better enterprise value.

What is Datical's relationship to Liqubase?

Cornett: There tends to be synergistic relationships between open source projects and strong open source companies. Linux  [The Linux Foundation] and Red Hat are a great example of that synergy as are MariaDB.org [MariaDB Foundation] and MariaDB.com [MariaDB Corp.] and we aim to do the same thing here. Liquibase is a very powerful open source solution, and Datical has built some monetization around that, and what we want to do is have more focus on integrating those two efforts than we have in the past. The founder and the key contributors of the Liquibase community are employees of Datical. We are focusing on tightening up this integration between this strong project and the success that Datical has already had in the enterprise market to the benefit of both the project and the company.

Datical recently announced new capabilities for Liquibase, including Targeted Rollback, available for users of Liquibase Pro.

Cornett: Part of the process of enhancing the relationship between business side and the community is making sure that resources can be put back into the community. One means of doing this is introducing functionality that layers on top of the Liquibase foundation.

Targeted rollback enables developers to make a number of changes in the database, and then go after a specific change that didn't work out the way they wanted. The targeted rollback capability is an example of enhanced functionality that Datical will continue to layer onto Liquibase to provide higher value to enterprises. And, of course, with that, there is a going to be a thoroughly tested and certified version of Liquibase.

Along with a rapidly evolving, innovative community that is continually making changes to the code, it is important that enterprises have a stable version that they trust to behave for their particular requirements. You will see down the road even more capabilities around administration, ease of use, and some limited indemnification.

This interview was conducted, edited, and condensed by Joyce Wells.



Newsletters

Subscribe to Big Data Quarterly E-Edition