Insights

将数据发送到 Power BI

概述

在本主题中，您可以了解如何利用 Insights 实时数据导出功能将数据发送到 Microsoft Power BI 并在其中生成报告。

备注：在配置 Microsoft Power BI 报告之前配置实时数据导出。

创建 Azure 流分析作业

在 Azure 中创建并配置资源以使用 Insights 数据。

登录到 portal.azure.com
创建 流分析作业。
1. 选择所需的资源组。
2. 选择 “创建资源”。
3. 选择“ 分析 ” > “流分析作业 ”，然后填写所需信息。
为 流分析作业添加输入。
1. 转到“ 流分析作业”。
2. 选择 “输入” > “添加输入”。
3. 选择所需的 Azure 事件中心并创建一个新的使用者组。
为流分析作业添加输出。
通过以下任一方式配置 作业查询 ：
- 获取流中的所有事件：
```
SELECT
    Timestamp,
    EventType,
    FolderName, 
    Job.ProcessName as ProcessName,
    Job.State as State,
    Job.RobotName as RobotName,
    Job.HostMachineName as HostMachineName,
    Job.[Key] as JobKey,
    Job.StartTime as JobStartTime,
    Job.EndTime as JobEndTime,
    Job.DisplayName as ProcessDisplayName
INTO
    [location-output]
FROM
    [location-input] TIMESTAMP BY [Timestamp]
WHERE
    EventType like '%job.%'SELECT
    Timestamp,
    EventType,
    FolderName, 
    Job.ProcessName as ProcessName,
    Job.State as State,
    Job.RobotName as RobotName,
    Job.HostMachineName as HostMachineName,
    Job.[Key] as JobKey,
    Job.StartTime as JobStartTime,
    Job.EndTime as JobEndTime,
    Job.DisplayName as ProcessDisplayName
INTO
    [location-output]
FROM
    [location-input] TIMESTAMP BY [Timestamp]
WHERE
    EventType like '%job.%'
```
- 投影和聚合输入。以下查询按时间和类型检索事件类型，从而提高了 Power BI 中大型数据集的性能：
```
SELECT
    System.Timestamp AS WindowEnd,
    [EventType],
    count(*) as EventCount
INTO
    [Location]
FROM
    [location]
GROUP BY TumblingWindow(Duration(second, 1)), EventTypeSELECT
    System.Timestamp AS WindowEnd,
    [EventType],
    count(*) as EventCount
INTO
    [Location]
FROM
    [location]
GROUP BY TumblingWindow(Duration(second, 1)), EventType
```
测试端到端数据流的查询，验证输入和输出的正确授权。
启动 流分析作业。
1. 转到“ 流分析作业概述”。
2. 单击“开始”。