Our series of ongoing specialist services projects continues. In this edition of Cloud Native Project of the Week: our latest project about optimizing the cloud native database of a consumer client.
A major direct sales giant has been testing a new branch management application backed by Couchbase that they are implementing gradually across several hundred branches nationwide. With the planned incremental release to branches, the IT team is challenged with optimizing cost and ensuring business continuity. Ease of operations and being able to scale on demand without a heavy compute commitment from the onset are other considerations.
The client’s Couchbase Data Platform implementation is currently on virtual machines in Azure. Biqmind proposed moving the database into Azure Kubernetes Service (AKS) to optimize compute resources and ensure the client is fully leveraging Couchbase’s proven manageability and performance at any scale.
We are running a technical spike on behalf of the client’s IT team to verify this recommendation. A technical spike is a focused effort to research, investigate, design and prototype solutions. It is a proven approach for reducing risks of a new technical approach, increasing story estimate reliability or understanding and refining a requirement.
Our specialists have:
- Identified KPIs for performance, cost, scalability, management and operations
- Set up and configured a new AKS cluster
- Installed Couchbase’s Autonomous Operator for Kubernetes, this component enables Couchbase to be run as a stateful database application in parallel to microservices applications on major Kubernetes distributions such as AKS
- Migrated the database onto the AKS version
- Run performance testing on both versions
With the project ongoing, the team will also be looking to implement data backup/restore and run performance testing on the application to compile metrics. The end goal is to provide a costing comparison based on estimated sizing and performance; and supporting recommendations for next steps.
Considering the full implementation footprint of this project, a technical spike saves time and provides a clear picture of risks and concrete actions required.