schema_version: 1
id: system-unavailability-during-incident-window
title: Service unavailability investigation
status: active
owner: claude-code
environment:
  profile: stable-premium
  cluster_id: '10109909546444159555'
  parent_id: '2013516099403907072'
  cluster_type: premium
  tidb_version: unknown
time_window:
  incident_at: '2026-03-31T02:35:00Z'
  allowed:
  - id: workspace_observed_window
    start: '2026-03-31T02:35:00Z'
    end: '2026-03-31T03:24:00Z'
prompt:
  text: 请帮我分析集群 10109909546444159555 在 2026-03-31T02:35:00Z 到 2026-03-31T03:24:00Z 这个时间窗内发生了什么。业务侧反馈该时段出现明显服务不可用、查询超时和 QPS 明显下跌。请判断最可能的根因，并给出关键证据链、影响判断和处理建议。
tags:
- service-unavailable
- qps-drop
- component-health
- request-timeout
case_tier: P0
cost_tier: low
followup:
  scheduled_max_followups: 0