What problem does it solve?
This Skill eliminates the guesswork and delays in PD (Placement Driver) incident response for TiDB clusters by providing a structured entry-point guide that quickly classifies issues and routes on-call engineers to the correct specialized diagnosis workflow, reducing mean time to resolution for both control-plane and data-plane PD problems.
Core Features & Use Cases
- Symptom-Based Routing: A clear mapping table that links observed PD incident symptoms (placement rule failures, region imbalance, RU throttling, ETCD capacity issues) to the correct specialized PD diagnosis skill.
- First Response Checklist: A step-by-step list of critical context to capture immediately when a PD incident occurs, including version records, recent operational changes, and impact scope.
- Triage Hints & Common Commands: Quick guidance to identify the most likely root cause category, plus reference for common PD CLI commands to run during initial investigation.
- Use Case: If your TiDB cluster is experiencing unexpected region eviction after a scatter operation, this Skill immediately directs you to the
pd-scheduling specialized skill instead of wasting time searching for relevant documentation.
Quick Start
Use the pd-oncall-handbook skill to triage your active PD incident and receive a direct link to the specialized PD diagnosis skill matching your observed symptoms.