Oracle Database 21c (21.3) comes with full production support for Oracle Sharding on Kubernetes and Docker!

We are excited to announce that customers can now take advantage of Kubernetes and Docker based deployment for sharded databases starting with Oracle Database 21c (21.3).

Oracle Sharding distributes segments of a data set across many databases (shards) on independent compute resources, on-premises, or in the cloud. It enables globally distributed, linearly scalable, multi-model databases. It requires no specialized hardware or software. Oracle Sharding does all of this while providing strong consistency, the full power of SQL, support for structured and unstructured data, and the Oracle Database ecosystem. Oracle Sharding meets data sovereignty requirements and supports applications that require low latency and high availability.

Oracle Sharding on Kubernetes

With many Oracle customers already using Kubernetes to deploy applications and databases, there is an increase in demand to manage large datasets with high-throughputs, and this is where Oracle Sharding can be very useful because it brings extreme scalability, fault isolation, and geographical distribution of data by distributing the data set across different shards.

Oracle Sharding on Kubernetes uses StatefulSet to provide stable, unique network identifiers and stable, persistent storage so you can create and manage your Oracle Sharding replica set natively in Kubernetes with Oracle supported helm and chart templates. Data is stored on a persistent volume so when a pod is recreated, all the data is still there. Some of the other benefits of running Oracle Sharding on Kubernetes are as follows:

• Quick deployments using pre-built configurations
• Rapid provisioning and de-provisioning of Oracle Sharding makes CI/CD integration easier
• Self-monitoring for pod-failures
• Elasticity (scale-in and scale out) without losing data
• Seamless lifecycle management including patching

The architecture diagram below provides an overview of Oracle Sharding deployment on Kubernetes for both on-premise and cloud. To get started with Oracle Sharding on Kubernetes, follow the instructions on GitHub.

Customers can also take advantage of running Oracle Sharding with Kubernetes by deploying the workload on Oracle Container Engine for Kubernetes (OKE) in Oracle Cloud which provides additional benefits like pre-built architecture patterns, pre-built images for Oracle Sharding as well as automation of Kubernetes, container, and database lifecycle management. To get started with Oracle Sharding on OKE, download the pre-built image from Oracle’s Container Registry .

Oracle Sharding on Docker

Docker provides a quick mechanism to deploy Oracle Sharding using an existing Oracle Single Instance Database Image. Some of the benefits of running Oracle Sharding on Docker are as follows:

• Simplified packaging includes Oracle Database and GDS to expedite delivery
• Automatic service startup based on input parameters (YAML or environment variables)
• Support for shard scale up and scale down based on input parameters

Oracle provides sample Docker build files to facilitate sharded database installation, configuration, and environment setup for DevOps users for all steps, including installing and configuring the Docker engine, creating a global service manager (shard director) and Oracle Database images, creating a network bridge, creating containers for the Oracle Sharding objects and shard director, and deploying the containers.

Find the instructions and downloads for sharded database deployment with Docker at GitHub repository

Additional Resources for Kubernetes and Docker support:

• Oracle Database Sharding on Kubernetes and Docker - Released Versions and Known Issues (Doc ID 2813935.1)
• Sharding Tools and deployment automation library at GitHub
• Oracle Sharding documentation for sharded database deployment with Terraform, Kubernetes, and Docker