Goal

• Allow specifying AWS Capacity blocks to ensure access to reserved EC2 instances requiring specialized hardware

Why is this important?

• AI/ML workloads capable instances get significant discounts when using capacity blocks

Scenarios

1. Hosted Cluster user on ROSA/self-managed-aws wants to run machine learning workloads with GPU acceleration

Acceptance Criteria

• Dev - Has a valid and documented implementation
• CI - MUST be running successfully with tests automated
• QE - covered in Polarion test plan and tests implemented

Dependencies (internal and external)

1. CAPA support

Previous Work (Optional):

1. CAPA support for CapacityResource
2. NodeLaunchTemplate example in EKS
3. Capacity blocks prerequisites

Open questions:

1. Should the capacity block be created on demand or just specified?

Done Checklist

• CI - CI is running, tests are automated and merged.
• DEV - Upstream documentation merged: <link to meaningful PR or GitHub Issue>
• QE - Test plans in Polarion: <link or reference to Polarion>
• QE - Automated tests merged: <link or reference to automated tests>
• DOC - Downstream documentation merged: <link to meaningful PR>