# Validation and Rollout

## Goal

The release target is not just feature completeness.
It also requires:

- Dedicated regression safety
- TiDB-X correctness for Premium / Essential / Starter
- operationally safe rollout

## Validation Focus

### 1. Dedicated Regression

- existing Dedicated metrics and labels must remain unchanged
- Dedicated create / update flows must keep working

### 2. TiDB-X Functional Validation

- Premium works for Prometheus and OTel
- Essential works for Prometheus and OTel
- Starter works for Prometheus and OTel
- output labels follow instance-level semantics
- datasource matcher behavior follows per-plan rules

### 3. Rollout Readiness

- `tidb-mgmt` connectivity is ready in target environments
- required bizTypes are registered
- real cluster verification exists for target plans and regions
- rollout evidence is recorded in an e2e or sign-off artifact

## Operational Safeguards Included in This Release

Besides the core TiDB-X feature, the release also includes operational safety changes that affect runtime behavior:

- pipeline backfill batching to reduce oversized catch-up windows
- topology hot-update related design work to reduce unnecessary pipeline restart pressure during scale changes

These are not the product-facing TiDB-X semantics themselves, but they matter for safe rollout and stable operation.

## Suggested Launch Checklist

1. Validate Dedicated baseline against historical output.
2. Validate Premium / Essential / Starter output labels and metric counts.
3. Validate service-plan-aware datasource routing in staging.
4. Validate `tidb-mgmt` lookup and Helm materialization.
5. Confirm rollout and rollback procedures are ready before production deploy.
6. Record results using `release/2026-03-15-e2e-test-report-template.md`.