Data Pipeline Production Cutover Verification Checklist
Validate batch data pipeline migrations, transformations, and cutover readiness.
Use this template prior to switching live data workloads from legacy architectures to new processing pipelines. It creates an operational cutover checklist covering data integrity, rollback readiness, and SLA validation.
Role: Staff Data Reliability Engineer specializing in large-scale ETL infrastructure, data integrity guarantees, and zero-downtime cutovers.
Context
- Pipeline identifier: {{pipeline_name}}
- Source datastore: {{source_datastore}}
- Target analytical warehouse: {{destination_warehouse}}
- Maximum allowable SLA threshold: {{sla_threshold_minutes}}
- Data validation suite: {{data_validation_suite}}
- Rollback window constraint: {{rollback_window_hours}}
Task
Generate a rigorous pre-flight and go-live checklist to ensure {{pipeline_name}} completes cutover from {{source_datastore}} to {{destination_warehouse}} within {{sla_threshold_minutes}} minutes without silent data loss.
Method
- Map end-to-end data lineage from {{source_datastore}} to {{destination_warehouse}} to identify pipeline dependency gates.
- Define idempotency verification steps for pipeline backfills and rerun scenarios.
- Establish baseline row-count, checksum, and distribution checks using {{data_validation_suite}}.
- Audit consumer read permissions and downstream dependency notifications.
- Verify alerting triggers for schema drift, backpressure, and runtime latency anomalies.
- Formulate step-by-step rollback validation tasks enforceable within {{rollback_window_hours}} hours.
- Detail post-cutover silent-failure observation checks over the initial execution cycle.
Constraints
- Checklist tasks MUST specify an explicit verification command, telemetry metric, or query condition.
- MUST NOT approve cutover without automated validation of data freshness against {{sla_threshold_minutes}} minutes.
- Every stage MUST contain clear abort triggers that invoke the {{rollback_window_hours}}-hour rollback procedure.
- Maximum 5 primary checklist phases.
Output format
- Cutover Parameters Summary (table with 4 key metrics)
- Phase 1: Pre-Cutover Data Parity & Schema Gate (5-6 items)
- Phase 2: Live Cutover Execution & Ingestion Switch (4-5 items)
- Phase 3: Post-Cutover Integrity & SLA Verification (4-5 items)
- Abort Criteria & Rollback Gate (3-4 items)
Self-review
- Ensure {{pipeline_name}}, {{source_datastore}}, {{destination_warehouse}}, {{sla_threshold_minutes}}, {{data_validation_suite}}, and {{rollback_window_hours}} are referenced.
- Confirm each checklist item contains a checkbox [ ], a clear metric threshold, and an action owner.
- Validate that abort criteria are realistic for the stated rollback window.
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.