-
Spike
-
Resolution: Unresolved
-
Undefined
-
None
-
None
-
None
-
None
-
False
-
-
False
-
None
-
None
-
None
1. Objective & Goal
Enable KEDA on IBM Power. The primary focus is to validate the provided multi-arch builds, verify installation, and identify technical gaps for General Availability (GA).
2. Target Outcomes (Deliverables) :
Image Inventory: A definitive list of container images required for Power.
Installation Verification: Confirm the operator installs successfully using the provided multi-arch build.
Test Execution Report: Run standard e2e tests (or document specific blockers preventing execution).
GA Requirements List: A documented list of blockers/requirements for General Availability (GA).
Upstream Integration Requirements: Defining the requirements for contributing the changes back to the core project to ensure long-term support.
All apply.
3. Key Dates
- Spike Start: 01/01/2026
- Spike Review: a.s.a.p.
- Target TP: n/a
- Target GA: 2Q 2026
4. Technical Considerations
- Target Architecture: IBM Power ppc64le
- Testing Scope: Identify e2e tests applicable to this architecture (excluding tests dependent on power-only services).
- Build Infrastructure: Assess upstream CI/CD pipeline capabilities (availability of emulation or native nodes).
- Base Image Compatibility: Verify that required Base OS container images (minimal/neutral variants) are available for this architecture.{}
5. Dependencies
| Dependency | Impact | Mitigating Action / Path to Green | Owner | Next Update |
| RH Autoscaling Team | CRITICAL: Source of provided builds. | Confirm delivery date of initial multi-arch images. | James Kryos | all / set |
| Peer Team (IBM Z) | Potential duplicated effort. | Sync with Z team to share findings on common build/test failures. | Vikas Singh | n/a |
| Hardware Access | Cannot verify builds/tests. | Confirm access to valid target environment (LPAR/Cluster). | Paul Bastide | n/a |
6. Risks
| Risk Description | Impact | Mitigating Action | Owner | Next Update |
| Upstream Rejection | Long-term maintenance burden. | Validate Upstream Contribution Requirements early in the spike. | [Name] | [Date] |
| Library Incompatibility | Scope Creep. | Scan dependencies for native binary requirements that may not compile on Power. | [Name] | [Date] |
7. MANAGEMENT & REVIEW (To be completed at Spike Conclusion)
- Feasibility Decision: [ GO / NO-GO / PIVOT ]
- Confidence Level (1-Low to 5-High): [ ]
- Est. Implementation Size (ROM): [ S / M / L / XL ]
- Next Steps:
- [ ] Created Jira Ticket for Implementation: [Link]
- [ ] Created Jira Ticket for Blocking Issues: [Link]
Original Spike Contents For Reference:
Custom Metrics AutoScaler (KEDA) for P AUTOSCALE-500
Add images
Support upstream
- relates to
-
MULTIARCH-5994 Support Custom Metrics AutoScaler (KEDA) for IBM Power
-
- New
-
- links to