What problem does it solve?
This Skill addresses critical TiKV cluster failures including region unavailability due to replica majority loss, unsafe recovery workflow failures, persistent orphan peers, and TiFlash-related recovery blockers that impact database uptime and data integrity for on-call engineering teams.
Core Features & Use Cases
- Unsafe Recovery Workflows: Provides version-specific guidance for both online (v6.1.0+) and manual unsafe recovery procedures, with clear recommendations for stable baseline versions.
- Edge Case Resolution: Includes workarounds for known recovery blockers such as learner node interference, TiFlash replica conflicts, and post-recovery orphan peer errors.
- Operational Safeguards: Outlines pre-recovery checks, data loss risk assessment steps, and post-recovery consistency validation to minimize unintended data loss.
- Use Case: If 2 of 3 TiKV replicas for a region are lost and the cluster reports "Region is unavailable", this Skill guides you through the correct recovery process for your cluster version, including steps to resolve stuck recovery flows caused by TiFlash replicas.
Quick Start
Use the tikv-recovery skill to walk you through the correct unsafe recovery workflow for your TiKV cluster version when regions are unavailable due to majority replica loss, including steps to clean up orphan peers and verify post-recovery cluster consistency.