tmf628-performance-manager

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

TMF628 Performance Analytics Manager

TMF628性能分析管理器

This skill provides a robust framework for analyzing network performance data using the TMF628 Performance Management API. It identifies network anomalies, calculates time-series averages, and ranks cell sites for field team dispatch based on threshold deviations.

本技能提供了一个基于TMF628性能管理API分析网络性能数据的稳健框架。它可以识别网络异常、计算时间序列平均值，并根据阈值偏差对蜂窝基站进行排序，以便现场团队调度。

Prerequisites

前提条件

MCP Server: Requires a TMF628 Performance Management v4 compatible server.
Runtime: Node.js for executing the analytics script.

MCP Server：需要兼容TMF628性能管理v4的服务器。
运行环境：Node.js，用于执行分析脚本。

MCP Server Requirements

MCP服务器要求

This skill requires an MCP server that provides TMF628 Performance Management capabilities.

Required tools:

Performance-Management-v4_listPerformanceIndicatorSpecification

Performance-Management-v4_listMeasurementCollectionJob

Performance-Management-v4_createMeasurementCollectionJob

Reference MCP Server: If you do not already have a locally configured MCP server for performance management, you can use:

URL:
```
https://tmf628.mcpgateway.online/mcp
```

本技能需要提供TMF628性能管理功能的MCP服务器。

所需工具：

Performance-Management-v4_listPerformanceIndicatorSpecification

、

Performance-Management-v4_listMeasurementCollectionJob

、

Performance-Management-v4_createMeasurementCollectionJob

。

参考MCP服务器：如果您尚未在本地配置用于性能管理的MCP服务器，可以使用：

URL：
```
https://tmf628.mcpgateway.online/mcp
```

Workflow

工作流程

1. Indicator & Threshold Discovery

1. 指标与阈值发现

When a user reports issues (e.g., "slow data" or "high utilization"), first discover the correct KPIs and their defined business thresholds.

List performance indicator specifications.
Map the user's issue to metric IDs (e.g.,
```
Cell Utilization
```
).
Extract Thresholds: Look at the
```
threshold
```
object in the specification. These values will be used for analysis.

当用户反馈问题（如“数据缓慢”或“高利用率”）时，首先找到正确的KPI及其定义的业务阈值。

列出性能指标规范。
将用户的问题映射到指标ID（例如
```
Cell Utilization
```
）。
提取阈值：查看规范中的
```
threshold
```
对象，这些值将用于分析。

2. Data Acquisition

2. 数据获取

List existing measurement collection jobs to find data for the target region.
If no recent job exists, create a new measurement collection job using standard TMF prefixes:
- ```
granularity
```
  :
```
g_15mn
```
- ```
reportingPeriod
```
  :
```
r_1h
```
Retrieve the results once the job is complete.

列出现有的测量收集任务，以查找目标区域的数据。
如果没有近期任务，使用标准TMF前缀创建新的测量收集任务：
- ```
granularity
```
  ：
```
g_15mn
```
- ```
reportingPeriod
```
  ：
```
r_1h
```
任务完成后获取结果。

3. Time-Series Analysis & Dispatch Ranking

3. 时间序列分析与调度排序

Process the performance data using the bundled script to identify violations over a specific time window.

Run the bundled
```
scripts/analyze-pm-data.js
```
script providing:
- The performance data JSON file.
- The threshold value to check against.
- The start and end hours for the analysis window (e.g.,
```
7
```
  and
```
9
```
  for a 7-9 AM peak).
The script filters data and ranks resources by deviation magnitude.

使用附带的脚本处理性能数据，识别特定时间窗口内的违规情况。

运行附带的
```
scripts/analyze-pm-data.js
```
脚本，需提供：
- 性能数据JSON文件。
- 要检查的阈值。
- 分析窗口的开始和结束小时数（例如，早高峰7点到9点使用
```
7
```
  和
```
9
```
  ）。
脚本会过滤数据，并根据偏差幅度对资源进行排序。

4. Present Actionable Report

4. 生成可执行报告

Format the output for operational use:

Hotspot Priority: List Resource IDs in order of failure severity.
Violation Details: For each resource, show the recorded
```
value
```
vs
```
threshold
```
and the
```
deviation
```
.
Summary: Total records scanned vs total violations found.

为运营使用格式化输出：

热点优先级：按故障严重程度列出资源ID。
违规详情：针对每个资源，显示记录的
```
value
```
与
```
threshold
```
以及
```
deviation
```
。
摘要：扫描的总记录数与发现的违规总数。

Platform Adaptation Notes

平台适配说明

Validation Constraints

验证约束

TMF628 is strict about enumeration values. Always use

g_

for granularity and

r_

for reporting periods.

TMF628对枚举值要求严格。粒度始终使用

g_

前缀，报告周期始终使用

r_

前缀。

Script Capabilities

脚本功能

The

analyze-pm-data.js

script handles filtering and ranking. It outputs a JSON object containing a summary and a sorted list of violations.

analyze-pm-data.js

脚本负责过滤和排序。它会输出包含摘要和违规排序列表的JSON对象。