The future of DBaaS on Kubernetes is clearly a huge area of interest for the database community. More than 200 people packed the conference room for an expert panel discussion on DBaaS and Kubernetes this year at KubeCon + CloudNativeCon North America. You can watch the full video of the panel here. EDB Vice President and Chief Architect Gabriele Bartolini joined a panel of cloud-native leaders to discuss challenges and opportunities in the future of databases on Kubernetes. Sharing the stage with Melissa Logan (Constantia), Sergey Pronin (Percona), and Deepthi Sigireddi (PlanetScale), Gabriele offered insights into the transformative potential of Kubernetes for database workloads, particularly PostgreSQL.
This session wasn’t just about trends—it was a deep dive into best practices and emerging solutions for database professionals embracing Kubernetes. The discussion highlighted how PostgreSQL, Kubernetes, and the evolving role of database administrators (DBAs) are converging to redefine the data landscape.
This blog dives deep into the technical learnings and thought leadership he contributed.
The Rise of Databases on Kubernetes
Kubernetes adoption for databases has surged in recent years, fueled by advancements in storage, networking, and automation. The panelists attributed this momentum to the evolution of Kubernetes-native capabilities, particularly in storage. The discussion highlighted how features like volume group snapshots have transformed day-two operations, such as backup and recovery, making them more reliable and efficient.
"Storage evolution has been pivotal," Gabriele noted. "From the introduction of local persistent volumes in 2019 to innovations in volume group snapshots, we’ve seen dramatic improvements in how we handle day-two operations like backup and recovery."
Because Postgres has an advanced native replication system you can do synchronous replication at the transaction level. This architecture allows high availability (HA) and disaster recovery while minimizing data loss. As Gabriele explains, “It’s why we now confidently recommend Kubernetes for mission-critical database workloads."
For PostgreSQL, Kubernetes offers unparalleled advantages, particularly in leveraging its native replication system. PostgreSQL’s synchronous replication capabilities allow organizations to achieve high availability (HA) and disaster recovery (DR) while minimizing downtime and data loss. By combining these strengths with Kubernetes' orchestration, companies can streamline their operations and reduce complexity.
Constantia.io CEO Melissa Logan pointed out that these advancements align with a larger trend: “For three years running, databases have been the number one workload on Kubernetes. Enterprises are trusting Kubernetes with their most critical data, which speaks to how far the technology has come.”
According to the 2024 Data on Kubernetes (DoK) Report, nearly half of organizations run 50% or more of their data workloads on Kubernetes in production. The report, which collects feedback from 150 technology leaders, shows Kubernetes is increasingly becoming a foundational infrastructure for AI/ML workloads while continuing to excel at traditional data workloads.
Best Practices for PostgreSQL on Kubernetes
Later in the discussion, the panel warned against a one-size-fits-all approach for running databases on Kubernetes, suggesting instead that tailoring database architectures to Kubernetes environments is key to success.
Key recommendations for PostgreSQL:
- Dedicated Nodes and Logical Placement: Allocating dedicated nodes with local storage ensures predictable performance and fault tolerance. Tools like taints, labels, and affinity rules allow PostgreSQL instances to be logically distributed across availability zones and nodes.
- Optimizing for Kubernetes' Strengths: PostgreSQL’s design should align with Kubernetes-native tools and practices. By relying on Kubernetes for observability, automation, and security, organizations can reduce complexity and enhance operational resilience.
- Embracing Microservice Databases: Bartolini highlighted an emerging trend where microservices and their associated databases coexist within the same namespace. This model empowers developers to manage databases directly, fostering agility and ownership.
Percona Product Manager Sergey Pronin stressed the importance of rethinking database management for Kubernetes: “The biggest mistake is treating Kubernetes as a lift-and-shift solution. It’s not about replicating what you did with VMs; it’s about embracing a cloud-native mindset to unlock new possibilities.”
The Evolving Role of DBAs in the Cloud-Native World
As Kubernetes adoption grows, DBAs are navigating a pivotal transition. Traditionally focused on managing infrastructure, they now find themselves bridging database expertise with Kubernetes fluency. Gabriele identified this shift as a critical opportunity: “DBAs need to adopt a T-shaped profile. They must deepen their knowledge of Kubernetes while leveraging their expertise in database systems like PostgreSQL. This shift mirrors the transition DBAs made when Linux became mainstream.”
Kubernetes democratizes access to powerful database capabilities because it empowers organizations to achieve cloud neutrality. With solutions like CloudNativePG, you can avoid vendor lock-in, maintain data sovereignty, and gain predictable performance—all while reducing costs. But transitioning databases to Kubernetes isn’t just a technical challenge; it’s a cultural one.
PlanetScale Software Engineer Deepthi Sigireddi echoed this sentiment, noting that automation is becoming essential for success. She explained that tools like Kubernetes operators empower developers to handle routine tasks, such as scaling and provisioning, reducing the workload for DBAs while ensuring consistency and reliability. “Self-service capabilities are key. With tools like Kubernetes operators, developers can manage their own databases, scaling resources as needed without human intervention,” says Deepthi.
Driving Innovation with CloudNativePG
CloudNativePG, the Kubernetes operator for PostgreSQL created and open-sourced by EDB, is a prime example of how open-source innovation is bridging PostgreSQL and Kubernetes. CloudNativePG supports use cases ranging from internal database-as-a-service (DBaaS) platforms to developer-managed microservice databases. It has been adopted by major organizations like IBM Cloud and Google Cloud, demonstrating its versatility and scalability.
Gabriele highlighted a compelling use case: startups leveraging Kubernetes and PostgreSQL to build disposable, repeatable environments for testing and development. “Kubernetes enables you to simulate multi-region clusters on a laptop,” he said. “This level of accessibility and flexibility is empowering the next wave of innovation.”
A Glimpse into the Future
The panel concluded with a forward-looking discussion on where databases and Kubernetes are headed. Melissa asked the panelists to share their predictions and aspirations for the future of data on Kubernetes, and what change will be needed to support the workloads of the future.
Gabriele emphasized that the future lies in tighter integration between these two ecosystems. At EDB, efforts are underway to enhance PostgreSQL’s extension management to align with Kubernetes’ dynamic capabilities, such as loading container images at runtime.
Deepthi predicted that managed database services will increasingly adopt Kubernetes as their backbone, citing its ability to handle the scale and complexity of thousands of databases. Pronin added that the continued maturity of Kubernetes and its ecosystem will lower barriers, enabling more organizations to adopt it confidently.
The panelists also highlighted the need to deepen the integration between databases and Kubernetes.
“At EDB, we’re working on enhancing PostgreSQL’s extension management to align with Kubernetes’ dynamic capabilities,” Gabriele explained. “The future lies in bridging these two ecosystems, enabling seamless updates and integrations at runtime.”
Deepthi forecasted a surge in managed database services powered by Kubernetes. “Kubernetes simplifies managing thousands of databases, which is why more providers are adopting it for their database platforms,” she said.
Sergey concluded with a call to embrace the maturing Kubernetes ecosystem, saying “As Kubernetes becomes more user-friendly, we’ll see tools emerge that make running databases even simpler—going beyond YAML manifests to deliver intuitive, automated experiences.”
A commitment to powering the databases of the future
The panel’s insights at KubeCon + CloudNativeCon 2024 underscored the exciting possibilities in the future of DBaaS on Kubernetes. EDB is proud to be contributing our knowledge and expertise to that conversation, as leaders in the PostgreSQL and Kubernetes ecosystems. As organizations continue to standardize on Kubernetes for both stateless and stateful workloads, EDB is shaping the tools and practices that will define the next generation of database management.
With CloudNativePG and EDB’s commitment to open-source innovation, the future of PostgreSQL in the cloud-native world looks brighter than ever. The panel underscored that databases on Kubernetes are no longer a bold experiment—they are a reality, driving cost savings, automation, and innovation.
Gabriele summed it up with an optimistic vision: “PostgreSQL and Kubernetes are evolving together, and the potential is immense. We’re at the forefront of enabling organizations to achieve cloud neutrality, reduce costs, and innovate with confidence.” You can read more insights from Gabriele about cloud neutral Postgres databases with Kubernetes and CloudNativePG in his latest blog for the Cloud Native Computing Foundation.
EDB is committed to advancing PostgreSQL and shaping the future of cloud-native data architectures. For organizations seeking to harness the power of PostgreSQL in the cloud-native world, the future is bright—and EDB is leading the charge.
