Goal

Perform a deep technical analysis of Masakari’s current dependency on Eventlet to assess feasibility, risks, and expected impact of a full removal. The outcome should enable stakeholders to clearly understand where and how Eventlet is used, and what is required to transition to native Python threading while preserving current behavior.

Acceptance Criteria

• A documented inventory of all Eventlet usage patterns across the Masakari codebase (monkey patching, greenthreads, WSGI server, etc.).

• A risk matrix that outlines migration complexity per usage area (low/medium/high).

• Behavior comparison between Eventlet and native threading for the key Masakari workflows.

• Recommendations for areas requiring redesign, refactoring, or alternative patterns.

• Clear feasibility assessment with go/no-go recommendation for full migration.

Open Questions

• Are there hidden transitive Eventlet dependencies beyond direct usage?

• Are there specific Masakari workloads where greenthreading behavior is critical and not safely replaced with threads?

• Do operators rely on any Eventlet-specific tuning that has no native equivalent?