CloudNativePG: Rethinking PostgreSQL Extensions in Kubernetes
Gabriele Fedi
EDB
CloudNativePG's Image Volume feature eliminates Postgres image rebuilds for extensions by bridging K8S Image Volumes with extension_control_path
In containerized PostgreSQL deployments, adding or updating extensions has always meant a painful cycle: rebuild your image, test it thoroughly, and redeploy your clusters with the new image. CloudNativePG's Image Volume feature streamlines this process—add extensions by simply declaring them in your CloudNativePG Cluster manifest, eliminating the image rebuild cycle and managing extensions independently from your database image.
This innovation works by connecting two powerful features: Kubernetes Image Volumes and PostgreSQL's
extension_control_path (introduced in PostgreSQL 18, with contributions from the CloudNativePG
community). By bridging these technologies, CloudNativePG brings a fundamentally new approach to
extension management in cloud-native environments.
In this talk, you'll learn:
- How CloudNativePG connects Kubernetes Image Volumes with PostgreSQL's extension_control_path to enable
dynamic extension loading
- Building standardized extension container images using postgres-extensions-containers
- Loading extensions on-demand into running PostgreSQL clusters without rebuilding images
- Managing extension lifecycles independently from your database images
- Real-world deployment patterns for popular extensions (pgvector, PostGIS, TimescaleDB, etc.)
- Security considerations and best practices for production environments
CloudNativePG: Rethinking PostgreSQL Extensions in Kubernetes
Gabriele Fedi
EDB
CloudNativePG's Image Volume feature eliminates Postgres image rebuilds for extensions by bridging K8S Image Volumes with extension_control_path
In containerized PostgreSQL deployments, adding or updating extensions has always meant a painful cycle: rebuild your image, test it thoroughly, and redeploy your clusters with the new image. CloudNativePG's Image Volume feature streamlines this process—add extensions by simply declaring them in your CloudNativePG Cluster manifest, eliminating the image rebuild cycle and managing extensions independently from your database image.
This innovation works by connecting two powerful features: Kubernetes Image Volumes and PostgreSQL's
extension_control_path (introduced in PostgreSQL 18, with contributions from the CloudNativePG
community). By bridging these technologies, CloudNativePG brings a fundamentally new approach to
extension management in cloud-native environments.
In this talk, you'll learn:
- How CloudNativePG connects Kubernetes Image Volumes with PostgreSQL's extension_control_path to enable
dynamic extension loading
- Building standardized extension container images using postgres-extensions-containers
- Loading extensions on-demand into running PostgreSQL clusters without rebuilding images
- Managing extension lifecycles independently from your database images
- Real-world deployment patterns for popular extensions (pgvector, PostGIS, TimescaleDB, etc.)
- Security considerations and best practices for production environments