---
schema_version: 1
id: read-pool-cpu-scope-compare-in-premium-cluster-during-latest-window
title: Single-node versus cluster-total unified read-pool CPU scope comparison
  in the primary premium cluster during the latest completed Beijing 15:10-16:00
  window
status: active
owner: codex
evaluation:
  mode: adhoc_evidence
environment:
  profile: stable-premium
  cluster_label: premium-main
  cluster_id: '10748116073341772391'
  cluster_type: premium
  tidb_version: v8.5.x
  parent_id: '2026948324023603200'
time_window:
  resolved_by_output: case_input
runtime:
  hooks:
  - id: prepare_case_input
    stage: before_acp
    kind: python_script
    path: runtime_hook.py
    output_id: case_input
    args:
    - prepare-case-input
    - "--cluster-id"
    - "{{environment.cluster_id}}"
    - "--metrics-cluster-id"
    - "{{environment.parent_id}}"
  - id: build_runtime_reference
    stage: before_judge
    kind: python_script
    path: runtime_hook.py
    output_id: reference_answer
    args:
    - build-reference
    - "--cluster-id"
    - "{{environment.cluster_id}}"
    - "--metrics-cluster-id"
    - "{{environment.parent_id}}"
    - "--start"
    - "{{runtime.case_input.start}}"
    - "--end"
    - "{{runtime.case_input.end}}"
prompt:
  text: |
    帮我精确看一下这个 Premium 集群（{{environment.cluster_id}}）在 {{runtime.case_input.start}} 到 {{runtime.case_input.end}} 这段时间里的 `tikv_unified_read_pool_cpu`。

    我只看最后结果。你可以先自行完成查询和计算，但在最终答案之前不要输出任何文本、解释、步骤、思考或铺垫。
    最终答案必须且只能是一个 `json` 代码块，字段必须严格为：
    `cluster_id`、`metrics_cluster_id`、`window`、`evidence_sufficient`、`peak_instance_node`、`peak_instance_read_pool_cpu_cores`、`cluster_total_peak_read_pool_cpu_cores`

    规则严格如下：
    1. `peak_instance_read_pool_cpu_cores` 指单个 TiKV 节点在该窗口内的最高 `tikv_unified_read_pool_cpu` 值；
    2. `cluster_total_peak_read_pool_cpu_cores` 指同一窗口内全 TiKV 节点 `tikv_unified_read_pool_cpu` 总和的最高值；
    3. 两个值都必须使用 CPU cores，不是百分比；
    4. `peak_instance_node` 必须输出 K8s pod 名；
    5. 所有 `*_cores` 字段保留 4 位小数；
    6. 不要输出额外字段，不要输出额外解释。

    JSON 结构如下：

    ```json
    {
      "cluster_id": "10748116073341772391",
      "metrics_cluster_id": "2026948324023603200",
      "window": {
        "start": "2026-01-01T07:10:00Z",
        "end": "2026-01-01T08:00:00Z"
      },
      "evidence_sufficient": true,
      "peak_instance_node": "db-xxxx-tikv-aaaaaa",
      "peak_instance_read_pool_cpu_cores": 5.8500,
      "cluster_total_peak_read_pool_cpu_cores": 23.0000
    }
    ```

    如果证据不足，请把 `evidence_sufficient` 设为 `false`，并把数值字段设为 `null`。
    除了这一个 `json` 代码块，不要输出任何其他内容。
tags:
- premium
- metric-scope
- read-pool-cpu
- structured-output
- adhoc-evidence
case_tier: P1
cost_tier: low
followup:
  scheduled_max_followups: 0
execution:
  acp_bash_timeout_s: 300