Azure 时序见解 Gen1 资源管理器Azure Time Series Insights Gen1 Explorer


这是一篇 Gen1 文章。This is a Gen1 article.

本文介绍 Azure 时序见解 Gen1 资源管理器 Web 应用的功能和选项。This article describes the features and options for the Azure Time Series Insights Gen1 Explorer web app. Azure 时序见解资源管理器演示了该服务提供的,可在你自己的环境中访问的强大数据可视化功能。The Azure Time Series Insights Explorer demonstrates the powerful data visualization capabilities provided by the service and can be accessed within your own environment.

Azure 时序见解是一种完全托管的分析、存储和可视化服务,让同时探索和分析数十亿 IoT 事件变得简单轻松。Azure Time Series Insights is a fully managed analytics, storage, and visualization service that makes it simple to explore and analyze billions of IoT events simultaneously. 它提供数据的全局视图,从而可以快速验证 IoT 解决方案并避免任务关键型设备出现代价高昂的故障时间。It gives you a global view of your data, which lets you quickly validate your IoT solution and avoid costly downtime to mission-critical devices. 可以近乎实时地发现隐藏的趋势、找出异常以及进行根本原因分析。You can discover hidden trends, spot anomalies, and conduct root-cause analyses in near real time.


有关演示环境的指导教程,请阅读 Azure 时序见解快速入门For a guided tour through the demonstration environment, read the Azure Time Series Insights quickstart.


在可以使用 Azure 时序见解资源管理器之前,必须:Before you can use Azure Time Series Insights Explorer, you must:

浏览和查询数据Explore and query data

将事件源连接到 Azure 时序见解环境的几分钟内,你可以浏览和查询时序数据。Within minutes of connecting your event source to your Azure Time Series Insights environment, you can explore and query your time series data.

  1. 若要开始,请在 Web 浏览器中打开 Azure 时序见解资源管理器To start, open the Azure Time Series Insights Explorer in your web browser. 在窗口左侧选择一个环境。On the left side of the window, select an environment. 有权访问的所有环境均按字母顺序列出。All environments that you have access to are listed in alphabetical order.

  2. 选择环境后,使用顶部的“从”和“到”配置,或者将其选中并拖到所需的时间跨度 。After you select an environment, either use the From and To configurations at the top, or select and drag over the timespan you want. 单击右上角的放大镜,或者右键单击所选时间跨度并选择“搜索” 。Select the magnifying glass in the upper-right corner, or right-click on the selected timespan and select Search.

  3. 还可以通过选择“自动打开” 按钮,每分钟自动刷新可用性。You also can refresh availability automatically every minute by selecting the Auto On button. “自动打开”按钮仅适用于可用性图表,而不适用于主要可视化效果的内容 。The Auto On button only applies to the availability chart, not the content of the main visualization.

  4. 单击 Azure 云图标会转到 Azure 门户中的环境。The Azure cloud icon takes you to your environment in the Azure portal.

    Azure 时序见解环境选择Azure Time Series Insights environment selection

  5. 接下来,会显示一个图表,该图表显示的是所选时间范围内所有事件的计数。Next, a chart that shows a count of all events during the selected timespan is displayed. 在此有许多控件:Here you have a number of controls:

    • 搜索词编辑器面板:术语空间是查询环境的位置。Terms Editor panel: The term space is where you query your environment. 它位于屏幕的左侧:It's found on the left side of the screen:
      • 度量值:此下拉列表显示所有数值列(双精度)。MEASURE: This drop-down list shows all numeric columns (Doubles).

      • 拆分依据:此下拉列表显示类别列(字符串)。SPLIT BY: This drop-down list shows categorical columns (Strings).

      • 可以启用梯级内插、显示最小值和最大值,以及在“度量值”旁边的控制面板中调整 Y 轴 。You can enable step interpolation, show minimum and maximum, and adjust the y-axis from the control panel next to MEASURE. 还可以调整显示的是数据的计数、平均值还是总和。You also can adjust whether data shown is a count, average, or sum of the data.

      • 最多可以添加要在同一个 X 轴上查看的五个搜索词。You can add up to five terms to view on the same x-axis. 选择“添加” 添加新的搜索词,或者使用“克隆此搜索词” 按钮添加现有搜索词的副本。Select Add to add a fresh term or use the Clone this term button to add a copy of an existing term.

        搜索词选择、筛选和查询面板Terms selection, filtering, and query panel

      • 谓词:使用谓词可以使用下表中列出的操作数集快速筛选事件。Predicate: Use the predicate to quickly filter your events by using the set of operands listed in the following table. 如果通过选择或单击执行搜索,谓词将基于该搜索自动更新。If you conduct a search by selecting or clicking, the predicate automatically updates based on that search. 支持的操作数类型包括:Supported operand types include:

        操作Operation 支持的类型Supported types 注释Notes
        <, >, <=, >=<, >, <=, >= 双精度DateTimeTimeSpanDouble, DateTime, TimeSpan
        =!=<>=, !=, <> 字符串布尔双精度DateTimeTimeSpanNULLString, Bool, Double, DateTime, TimeSpan, NULL
        ININ 字符串布尔双精度DateTimeTimeSpanNULLString, Bool, Double, DateTime, TimeSpan, NULL 所有操作数应为同一类型或者是 NULL 常数。All operands should be of the same type or be NULL constant.
        HASHAS 字符串String 右侧只允许使用常量字符串文本。Only constant string literals are allowed at the right side. 不允许空字符串和 NULLEmpty string and NULL aren't allowed.
      • 示例查询Example queries

        Gen1 查询示例Example Gen1 queries

  6. 可以使用“间隔大小” 滑块工具在相同时间跨度内放大和缩小间隔。You can use the Interval Size slider tool to zoom in and out of intervals over the same timespan. 使用该滑块可以更精确地控制在大量时间切片之间的移动,显示平滑的趋势直至小到毫秒的切片,从而可以显示并分析数据更精细、分辨率更高的剪切片段。The slider provides more precise control of movement between large slices of time that show smooth trends down to slices as small as the millisecond, which allow you to display and analyze granular, high-resolution cuts of your data. 滑块的默认起点设置为从你的选择中获得的最佳数据视图,以均衡分辨率、查询速度和粒度。The slider's default starting point is set as the most optimal view of the data from your selection to balance resolution, query speed, and granularity.

  7. 使用“时间画笔”工具可以轻松从一个时间跨度导航到另一个时间跨度。 The Time brush tool makes it easy to navigate from one timespan to another.

  8. 选择“保存” 图标可以保存当前查询,并与环境的其他用户共享该查询。Select the Save icon to save your current query and share it with other users of the environment. 选择“打开” 图标时,可以查看所有已保存的查询和有权访问的环境中的其他用户的任何共享查询。When you select the Open icon, you can review all of your saved queries and any shared queries of other users in environments you have access to.


可视化数据Visualize data

  1. 使用“透视视图” 工具可以查看最多四个唯一查询的同步视图。Use the Perspective View tool for a simultaneous view of up to four unique queries. “透视视图”按钮位于图表右上角 。The Perspective View button is in the upper-right corner of the chart.

    选择要添加到透视窗格中的查询Select queries to add to the perspective pane

  2. 查看图表可以直观浏览数据;使用“图表”工具 :View a chart to visually explore your data, and use the Chart tools:

    • 选择单击 特定的时间跨度或单个数据系列。Select or click a specific timespan or a single data series.
    • 在选择的时间跨度内可以缩放或浏览事件。Within a timespan selection, you can zoom or explore events.
    • 在数据系列内,可以按另一列拆分系列、将系列作为新术语添加、只显示已选系列、排除已选系列、固定该系列或从已选系列中浏览事件。Within a data series, you can split the series by another column, add the series as a new term, show only the selected series, exclude the selected series, ping that series, or explore events from the selected series.
    • 在图表左侧的筛选器区域中,可以查看显示的所有数据系列并按值或名称重新排序。In the filter area to the left of the chart, you can review all displayed data series and reorder by value or name. 还可以查看所有数据系列或者任意已固定或取消固定的系列。You also can view all data series or any pinned or unpinned series. 可以选择一个数据系列并按另一列拆分该系列、将系列作为新搜索词添加、只显示已选系列、排除已选系列、固定该系列或从已选系列中浏览事件。You can select a single data series and split the series by another column, add the series as a new term, show only the selected series, exclude the selected series, pin that series, or explore events from the selected series.
    • 同时查看多个术语时,可以堆叠、取消堆叠、查看有关数据系列的其他数据,以及对所有搜索词使用同一个 Y 轴。When you view multiple terms simultaneously, you can stack, unstack, review additional data about a data series, and use the same y-axis across all terms. 请使用图表右上角的按钮。Use the buttons in the upper-right corner of the chart.

    图表工具右上角选项设置Chart tool upper right corner option settings

  3. 使用 热度地图 可以快速发现给定查询中的唯一或异常数据系列。Use the heatmap to quickly spot unique or anomalous data series in a given query. 仅有一个搜索术语能可视化为热度地图。Only one search term can be visualized as a heatmap.

    时序见解资源管理器热度地图图表](./media/time-series-insights-explorer/tsi-ga-example-heatmap-charting.png#lightbox)Time Series Insights explorer heatmap charting](./media/time-series-insights-explorer/tsi-ga-example-heatmap-charting.png#lightbox)

  4. 通过选择或右键单击浏览事件时,会显示“事件”面板 。When you explore events by selecting or right-clicking, the EVENTS panel is made available. 在这里,可以查看所有原始事件并可将事件导出为 JSON 或 CSV 文件。Here, you can review all of your raw events and export your events as JSON or CSV files. Azure 时序见解存储所有原始数据。Azure Time Series Insights stores all raw data.


  5. 浏览事件以公开模式和列统计信息之后,选择“统计信息” 选项卡。Select the STATS tab after you explore events to expose patterns and column stats.

    • 模式:此功能主动显示所选数据区域中最具统计意义的模式。Patterns: This feature proactively surfaces the most statistically significant patterns in a selected data region. 无需查看数千个事件来了解最需要花费时间和精力的模式。You don't have to look at thousands of events to understand what patterns require the most time and energy. 你可以使用 Azure 时序见解直接跳转到这些具有重大统计意义的模式继续进行分析。With Azure Time Series Insights, you can jump directly into these statistically significant patterns to continue conducting an analysis. 此功能也有助于对历史数据进行事后调查。This feature is also helpful for post-mortem investigations into historical data.

    • 列统计信息:列统计信息提供图表和表格,细分选定时间范围内所选数据系列的每列中的数据。Column Stats: Column stats provide charts and tables that break down data from each column of the selected data series over the selected timespan.

      “统计信息”列图表和选项STATS column charting and options

现在,你已了解 Azure 时序见解资源管理器 Web 应用中可用的关键功能、配置设置和显示选项。Now you've learned about the key features, configuration settings, and display options that are available in the Azure Time Series Insights explorer web app.

后续步骤Next steps