Task

Pre-install operators on cluster pool nodes so that PR and nightly jobs skip runtime operator installation. Saves ~2-4 min per job and reduces flake surface area.

Background

Clusters are acquired from RHDH pools (RHDH-4-16/4-17/4-18/4-19-US-EAST-2, all AWS us-east-2). Configured in openshift/release. Clusters are hot but NOT preloaded with operators.

Pool-specific pre-warm:

• PR pools: Crunchy Postgres only (Pipelines not needed per QW-2)
• Nightly pools: Full stack (Pipelines + Crunchy + Serverless + SonataFlow)

Pipeline scripts must detect pre-warmed state via ConfigMap and skip redundant installs, with fallback to runtime install if detection fails.

Dependencies and Blockers

• Depends on QW-2 findings (which operators per pool type)
• Requires coordination with openshift/release maintainers for pool configuration

Acceptance Criteria

• PR cluster pool pre-warms with Crunchy Postgres
• Nightly pools pre-warm with full operator stack
• Measured savings: 2-4 min per job
• Drift detection: verify operator health before relying on pre-warm
• Fallback: runtime install if pre-warm not detected