Kubernetes wouldn’t be the backbone of stateful database applications if not for the work of talented database developers like Leonardo Cecchi. As one of the creators of CloudNativePG and a valued engineer at EDB, Leonardo was recently recognized for his significant contributions to tech storage and invited to join the Kubernetes Container Storage Interface (CSI) team.
Cloud Native Storage Tech Lead at VMware Xing Yang, and EDB Cloud Native VP Gabriele Bartolini, nominated and sponsored Leonardo for CSI team membership. “Leonardo's innovative work as one of the creators of CloudNativePG and his relentless research on integrating Kubernetes and PostgreSQL serve as a beacon of inspiration for us all,” says Gabriele.
Kubernetes has come a long way from its origins as an orchestration system for stateless workloads that weren’t focused on storage. Today, databases are a critical part of every workload, and the CloudNativePG operator that Gabriele and Leonardo and their colleagues have created and contributed to ensures that Postgres databases can successfully run in Kubernetes. Through the Native Computing Foundation (CNCF), they’re getting the word out and making cloud universal and sustainable.
CloudNativePG’s popularity has continued to increase with Leonardo’s work on volume snapshots and volume group snapshots. With the recent introduction of native support for Kubernetes volume snapshot backup and recovery, CloudNativePG is capable of restoring a 4.5 TB database in just minutes.
Previously, backup in traditional infrastructures had been complicated, because many different storage systems work with a different API. This means every storage system that customers are interested in has to be supported, which is virtually impossible. This is why most Postgres solutions aren’t based on snapshots.
But by relying on the standard in Kubernetes, Leonardo and his team enabled snapshots of all the storage systems implementing it. “It’s like a common interface,” says Leonardo. “Once you have it, every application can take a snapshot simply relying on this common instance instead of relying on that particular implementation of that storage. This is not only helping application developers, but providers as well, because it facilitates customers moving from one solution to another,” he says.
When we interviewed Leo for this article, he emphasized how important it is to strike a balance when using Postgres. “You should take, but you should also give,” he says. “This is what EDB is doing, contributing developers’ time to make Postgres great. And this is what I am doing with Kubernetes, and why I started contributing,” he says.
Leo points out that the recognition he received for tech storage in Kubernetes isn’t just about him, but about the relationship between the database and storage communities to ensure compatibility and seamless integration.
“We should be proud of this achievement,” he says, “not because of my work, but because we have created an ecosystem, and we are powering each other.”