快速入门:使用客户端库或 REST API 自定义解决方案

重要

从 2023 年 9 月 20 日开始,将无法创建新的 Azure 指标顾问资源。 指标顾问服务将于 2026 年 10 月 1 日停用。

开始使用指标顾问 REST API 或客户端库。 请按照以下步骤安装程序包并试用基本任务的示例代码。

使用指标顾问执行以下操作:

  • 从数据源添加数据馈送
  • 检查引入状态
  • 配置检测和警报
  • 查询异常情况检测结果
  • 诊断异常

参考文档 | 库源代码 | 包 (NuGet) | 示例

重要

Azure 建议使用最安全的可用身份验证流。 本文所述的某些身份验证流需要极高度地信任应用程序,并且附带了其他更安全的流中不存在的风险。 请仅在无法使用其他更安全的流(例如托管标识)时才使用此流。

先决条件

提示

  • 可以在 GitHub 上找到 .NET 指标顾问示例。
  • 指标顾问资源可能需要 10 到 30 分钟才能部署一个服务实例供你使用。 部署成功后,选择“转到资源”。 部署完成后,可以通过 Web 门户和 REST API 这两种方式开始使用指标顾问实例。
  • 可以在 Azure 门户的资源“概述”部分中找到 REST API 的 URL。 它将如下所示:
  • https://<instance-name>.cognitiveservices.azure.cn/

设置

安装客户端库

创建新项目后,右键单击“解决方案资源管理器”中的项目解决方案,然后选择“管理 NuGet 包”,以安装客户端库。 在打开的包管理器中,选择“浏览”,选中“包括预发行版”并搜索 Azure.AI.MetricsAdvisor 选择版本 1.0.0,然后选择“安装”。

在控制台窗口(例如 cmd、PowerShell 或 Bash)中,使用 dotnet new 命令创建名为 metrics-advisor-quickstart 的新控制台应用。 此命令将创建包含单个源文件的简单“Hello World”C# 项目:program.cs

dotnet new console -n metrics-advisor-quickstart

将目录更改为新创建的应用文件夹。 可使用以下代码生成应用程序:

dotnet build

生成输出不应包含警告或错误。

...
Build succeeded.
 0 Warning(s)
 0 Error(s)
...

如果你使用的 IDE 而不是 Visual Studio,则可通过以下命令安装适用于 .NET 的指标顾问客户端库:

dotnet add package Azure.AI.MetricsAdvisor --version 1.1.0

环境变量

若要成功对异常检测器服务发出调用,需要使用以下值:

变量名称
METRICS_ADVISOR_ENDPOINT 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到此值。 终结点示例:https://YOUR_RESOURCE_NAME.cognitiveservices.azure.cn/
METRICS_ADVISOR_KEY 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到键值。 可以使用 KEY1KEY2
METRICS_ADVISOR_API_KEY 指标顾问门户检查资源时,可以在“设置”>“API 密钥”下找到密钥值。 可以使用 KEY1KEY2
SQL_CONNECTION_STRING 本快速入门要求你拥有自己的 SQL 数据库 + 连接字符串。 示例连接字符串类似于以下示例:Data Source=<Server>;Initial Catalog=<db-name>;User ID=<user-name>;Password=<password>有关构造 SQL 连接字符串的详细信息,请参阅 SQL 文档
SQL_QUERY 特定于数据集的唯一查询。

创建环境变量

为密钥和终结点创建和分配持久环境变量。

重要

如果使用 API 密钥,请将其安全地存储在某个其他位置,例如 Azure Key Vault 中。 请不要直接在代码中包含 API 密钥,并且切勿公开发布该密钥。

有关 Azure AI 服务安全性的详细信息,请参阅对 Azure AI 服务的请求进行身份验证

setx METRICS_ADVISOR_ENDPOINT "REPLACE_WITH_YOUR_ENDPOINT_HERE" 
setx METRICS_ADVISOR_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx METRICS_ADVISOR_API_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx SQL_CONNECTION_STRING "REPLACE_WITH_YOUR_UNIQUE_SQL_CONNECTION_STRING" 
setx SQL_QUERY "REPLACE_WITH_YOUR_UNIQUE_SQL_QUERY_BASED_ON_THE_UNDERLYING_STRUCTURE_OF_YOUR_DATA" 

创建应用程序

编辑 program.cs 文件并替换为以下内容:

// Copyright (c) Microsoft Corporation. All rights reserved.
// Licensed under the MIT License.

using System;
using System.Threading.Tasks;
using Azure.AI.MetricsAdvisor.Administration;
using Azure.AI.MetricsAdvisor.Models;
using Azure.AI.MetricsAdvisor.Tests;
using Azure.Core.TestFramework;
using NUnit.Framework;
using static System.Environment;

namespace Azure.AI.MetricsAdvisor.Samples
{
    [LiveOnly]
    public partial class MetricsAdvisorSamples : MetricsAdvisorTestEnvironment
    {
        [Test]
        public async Task CreateAndDeleteDataFeedAsync()
        {
            string endpoint =  GetEnvironmentVariable("METRICS_ADVISOR_ENDPOINT");
            string subscriptionKey = GetEnvironmentVariable("METRICS_ADVISOR_KEY");
            string apiKey = GetEnvironmentVariable("METRICS_ADVISOR_API_KEY");
            var credential = new MetricsAdvisorKeyCredential(subscriptionKey, apiKey);

            var adminClient = new MetricsAdvisorAdministrationClient(new Uri(endpoint), credential);

            #region Snippet:CreateDataFeedAsync
#if SNIPPET
            string sqlServerConnectionString = GetEnvironmentVariable("SQL_CONNECTION_STRING");
            string sqlServerQuery = GetEnvironmentVariable("SQL_QUERY");
#else
            string sqlServerConnectionString = SqlServerConnectionString;
            string sqlServerQuery = SqlServerQuery;
#endif

            var dataFeed = new DataFeed();

#if SNIPPET
            dataFeed.Name = "<dataFeedName>";
#else
            dataFeed.Name = GetUniqueName();
#endif
            dataFeed.DataSource = new SqlServerDataFeedSource(sqlServerConnectionString, sqlServerQuery);
            dataFeed.Granularity = new DataFeedGranularity(DataFeedGranularityType.Daily);

            dataFeed.Schema = new DataFeedSchema();
            dataFeed.Schema.MetricColumns.Add(new DataFeedMetric("cost"));
            dataFeed.Schema.MetricColumns.Add(new DataFeedMetric("revenue"));
            dataFeed.Schema.DimensionColumns.Add(new DataFeedDimension("category"));
            dataFeed.Schema.DimensionColumns.Add(new DataFeedDimension("region"));

            dataFeed.IngestionSettings = new DataFeedIngestionSettings(DateTimeOffset.Parse("2020-01-01T00:00:00Z"));

            Response<DataFeed> response = await adminClient.CreateDataFeedAsync(dataFeed);

            DataFeed createdDataFeed = response.Value;

            Console.WriteLine($"Data feed ID: {createdDataFeed.Id}");
            Console.WriteLine($"Data feed status: {createdDataFeed.Status.Value}");
            Console.WriteLine($"Data feed created time: {createdDataFeed.CreatedOn.Value}");

            Console.WriteLine($"Data feed administrators:");
            foreach (string admin in createdDataFeed.Administrators)
            {
                Console.WriteLine($" - {admin}");
            }

            Console.WriteLine($"Metric IDs:");
            foreach (DataFeedMetric metric in createdDataFeed.Schema.MetricColumns)
            {
                Console.WriteLine($" - {metric.Name}: {metric.Id}");
            }

            Console.WriteLine($"Dimensions:");
            foreach (DataFeedDimension dimension in createdDataFeed.Schema.DimensionColumns)
            {
                Console.WriteLine($" - {dimension.Name}");
            }
            #endregion

            // Delete the created data feed to clean up the Metrics Advisor resource. Do not perform this
            // step if you intend to keep using the data feed.

            await adminClient.DeleteDataFeedAsync(createdDataFeed.Id);
        }

        [Test]
        public async Task GetDataFeedAsync()
        {
            string endpoint = GetEnvironmentVariable("METRICS_ADVISOR_ENDPOINT");
            string subscriptionKey = GetEnvironmentVariable("METRICS_ADVISOR_KEY");
            string apiKey = GetEnvironmentVariable("METRICS_ADVISOR_API_KEY");
            var credential = new MetricsAdvisorKeyCredential(subscriptionKey, apiKey);

            var adminClient = new MetricsAdvisorAdministrationClient(new Uri(endpoint), credential);

            string dataFeedId = DataFeedId;

            Response<DataFeed> response = await adminClient.GetDataFeedAsync(dataFeedId);

            DataFeed dataFeed = response.Value;

            Console.WriteLine($"Data feed status: {dataFeed.Status.Value}");
            Console.WriteLine($"Data feed created time: {dataFeed.CreatedOn.Value}");

            Console.WriteLine($"Data feed administrators:");
            foreach (string admin in dataFeed.Administrators)
            {
                Console.WriteLine($" - {admin}");
            }

            Console.WriteLine($"Metric IDs:");
            foreach (DataFeedMetric metric in dataFeed.Schema.MetricColumns)
            {
                Console.WriteLine($" - {metric.Name}: {metric.Id}");
            }

            Console.WriteLine($"Dimensions:");
            foreach (DataFeedDimension dimension in dataFeed.Schema.DimensionColumns)
            {
                Console.WriteLine($" - {dimension.Name}");
            }
        }

        [Test]
        public async Task UpdateDataFeedAsync()
        {
            string endpoint = GetEnvironmentVariable("METRICS_ADVISOR_ENDPOINT");
            string subscriptionKey = GetEnvironmentVariable("METRICS_ADVISOR_KEY");
            string apiKey = GetEnvironmentVariable("METRICS_ADVISOR_API_KEY");
            var credential = new MetricsAdvisorKeyCredential(subscriptionKey, apiKey);

            var adminClient = new MetricsAdvisorAdministrationClient(new Uri(endpoint), credential);

            string dataFeedId = DataFeedId;

            Response<DataFeed> response = await adminClient.GetDataFeedAsync(dataFeedId);
            DataFeed dataFeed = response.Value;

            string originalDescription = dataFeed.Description;
            dataFeed.Description = "This description was generated by a sample.";

            // Some properties, such as IngestionStartOffset, can be reset to their default value
            // when set to null during an Update operation. Check the API documentation to verify
            // when a property supports this feature.

            TimeSpan? originalStartOffset = dataFeed.IngestionSettings.IngestionStartOffset;
            dataFeed.IngestionSettings.IngestionStartOffset = null;

            response = await adminClient.UpdateDataFeedAsync(dataFeed);
            DataFeed updatedDataFeed = response.Value;

            Console.WriteLine($"Updated description: {updatedDataFeed.Description}");
            Console.WriteLine($"Updated ingestion start offset: {updatedDataFeed.IngestionSettings.IngestionStartOffset}");

            // Undo the changes to leave the data feed unaltered. Skip this step if you intend to keep
            // the changes.

            dataFeed.Description = originalDescription;
            dataFeed.IngestionSettings.IngestionStartOffset = originalStartOffset;

            await adminClient.UpdateDataFeedAsync(dataFeed);
        }

        [Test]
        public async Task GetDataFeedsAsync()
        {
            string endpoint = GetEnvironmentVariable("METRICS_ADVISOR_ENDPOINT");
            string subscriptionKey = GetEnvironmentVariable("METRICS_ADVISOR_KEY");
            string apiKey = GetEnvironmentVariable("METRICS_ADVISOR_API_KEY");
            var credential = new MetricsAdvisorKeyCredential(subscriptionKey, apiKey);

            var adminClient = new MetricsAdvisorAdministrationClient(new Uri(endpoint), credential);

            var filter = new DataFeedFilter()
            {
                Status = DataFeedStatus.Active,
                GranularityType = DataFeedGranularityType.Daily
            };
            var options = new GetDataFeedsOptions()
            {
                Filter = filter,
                MaxPageSize = 5
            };

            int dataFeedCount = 0;

            await foreach (DataFeed dataFeed in adminClient.GetDataFeedsAsync(options))
            {
                Console.WriteLine($"Data feed ID: {dataFeed.Id}");
                Console.WriteLine($"Name: {dataFeed.Name}");
                Console.WriteLine($"Description: {dataFeed.Description}");
                Console.WriteLine();

                // Print at most 5 data feeds.
                if (++dataFeedCount >= 5)
                {
                    break;
                }
            }
        }
    }
}

运行应用程序

从应用程序目录使用 dotnet run 命令运行应用程序。

dotnet run

参考文档 | 库源代码 | 项目 (Maven) | 示例

重要

Azure 建议使用最安全的可用身份验证流。 本文所述的某些身份验证流需要极高度地信任应用程序,并且附带了其他更安全的流中不存在的风险。 请仅在无法使用其他更安全的流(例如托管标识)时才使用此流。

先决条件

提示

  • 可以在 GitHub 上找到 Java 指标顾问示例。
  • 指标顾问资源可能需要 10 到 30 分钟才能部署一个服务实例供你使用。 部署成功后,选择“转到资源”。 部署完成后,可以通过 Web 门户和 REST API 这两种方式开始使用指标顾问实例。
  • 可以在 Azure 门户的资源“概述”部分中找到 REST API 的 URL。 它将如下所示:
    • https://<instance-name>.cognitiveservices.azure.cn/

设置

创建新的 Gradle 项目

本快速入门使用 Gradle 依赖项管理器。 可在 Maven 中央存储库中找到有关客户端库的详细信息。

在控制台窗口(例如 cmd、PowerShell 或 Bash)中,为应用创建一个新目录并导航到该目录。

mkdir myapp && cd myapp

从工作目录运行 gradle init 命令。 此命令将创建 Gradle 的基本生成文件,包括 build.gradle.kts,在运行时将使用该文件创建并配置应用程序。

gradle init --type basic

当提示你选择一个 DSL 时,选择 Kotlin

安装客户端库

找到 build.gradle.kts,并使用喜好的 IDE 或文本编辑器将其打开。 然后将以下生成配置复制到其中。 请确保包含项目依赖项。

dependencies {
    compile("com.azure:azure-ai-metricsadvisor:1.1.8")
}

创建 Java 文件

为示例应用创建一个文件夹。 在工作目录中运行以下命令:

mkdir -p src/main/java

导航到新文件夹,并创建名为 MetricsAdvisorQuickstarts.java 的文件。 在喜好的编辑器或 IDE 中打开该文件并添加以下 import 语句:

提示

想要立即查看整个快速入门代码文件? 可以在 GitHub 上找到它,其中包含此快速入门中的代码示例。

环境变量

若要成功对异常检测器服务发出调用,需要使用以下值:

变量名称
METRICS_ADVISOR_ENDPOINT 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到此值。 终结点示例:https://YOUR_RESOURCE_NAME.cognitiveservices.azure.cn/
METRICS_ADVISOR_KEY 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到键值。 可以使用 KEY1KEY2
METRICS_ADVISOR_API_KEY 指标顾问门户检查资源时,可以在“设置”>“API 密钥”下找到密钥值。 可以使用 KEY1KEY2
SQL_CONNECTION_STRING 本快速入门要求你拥有自己的 SQL 数据库 + 连接字符串。 示例连接字符串类似于以下示例:Data Source=<Server>;Initial Catalog=<db-name>;User ID=<user-name>;Password=<password>有关构造 SQL 连接字符串的详细信息,请参阅 SQL 文档
SQL_QUERY 特定于数据集的唯一查询。

创建环境变量

为密钥和终结点创建和分配持久环境变量。

重要

如果使用 API 密钥,请将其安全地存储在某个其他位置,例如 Azure Key Vault 中。 请不要直接在代码中包含 API 密钥,并且切勿公开发布该密钥。

有关 Azure AI 服务安全性的详细信息,请参阅对 Azure AI 服务的请求进行身份验证

setx METRICS_ADVISOR_ENDPOINT "REPLACE_WITH_YOUR_ENDPOINT_HERE" 
setx METRICS_ADVISOR_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx METRICS_ADVISOR_API_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx SQL_CONNECTION_STRING "REPLACE_WITH_YOUR_UNIQUE_SQL_CONNECTION_STRING" 
setx SQL_QUERY "REPLACE_WITH_YOUR_UNIQUE_SQL_QUERY_BASED_ON_THE_UNDERLYING_STRUCTURE_OF_YOUR_DATA" 

创建应用程序

将 .java 文件的内容替换为以下内容:

package com.azure.ai.metricsadvisor.administration;

import com.azure.ai.metricsadvisor.administration.models.AzureAppInsightsDataFeedSource;
import com.azure.ai.metricsadvisor.administration.models.DataFeed;
import com.azure.ai.metricsadvisor.administration.models.DataFeedDimension;
import com.azure.ai.metricsadvisor.administration.models.DataFeedGranularity;
import com.azure.ai.metricsadvisor.administration.models.DataFeedGranularityType;
import com.azure.ai.metricsadvisor.administration.models.DataFeedIngestionSettings;
import com.azure.ai.metricsadvisor.administration.models.DataFeedMetric;
import com.azure.ai.metricsadvisor.administration.models.DataFeedOptions;
import com.azure.ai.metricsadvisor.administration.models.DataFeedSchema;
import com.azure.ai.metricsadvisor.administration.models.DataFeedSourceType;
import com.azure.ai.metricsadvisor.models.MetricsAdvisorKeyCredential;

import java.time.OffsetDateTime;
import java.util.Arrays;
import java.util.Collections;

/**
 * Sample demonstrates how to create, get, update, delete and list datafeed.
 */
public class DatafeedSample {
    private static String subscription_key = System.getenv("METRICS_ADVISOR_KEY");
    private static String api_key = System.getenv("METRICS_ADVISOR_API_KEY");
    private static String endpoint = System.getenv("METRICS_ADVISOR_ENDPOINT");
    private static String connection_string = System.getenv("SQL_CONNECTION_STRING");
    private static String sql_query = System.getenv("SQL_QUERY");

    public static void main(String[] args) {
        final MetricsAdvisorAdministrationClient advisorAdministrationClient =
            new MetricsAdvisorAdministrationClientBuilder()
                .endpoint("https://{endpoint}.cognitiveservices.azure.cn/")
                .credential(new MetricsAdvisorKeyCredential("subscription_key", "api_key"))
                .buildClient();

        // Create Data feed
DataFeed dataFeed = new DataFeed()
    .setName("dataFeedName")
    .setSource(new MySqlDataFeedSource(connection_string, sql_query))
    .setGranularity(new DataFeedGranularity().setGranularityType(DataFeedGranularityType.DAILY))
    .setSchema(new DataFeedSchema(
        Arrays.asList(
            new DataFeedMetric("cost"),
            new DataFeedMetric("revenue")
        )).setDimensions(
        Arrays.asList(
            new DataFeedDimension("city"),
            new DataFeedDimension("category")
        ))
    )
    .setIngestionSettings(new DataFeedIngestionSettings(OffsetDateTime.parse("2020-01-01T00:00:00Z")))
    .setOptions(new DataFeedOptions()
        .setDescription("data feed description")
        .setRollupSettings(new DataFeedRollupSettings()
            .setRollupType(DataFeedRollupType.AUTO_ROLLUP)));
final DataFeed createdSqlDataFeed = metricsAdvisorAdminClient.createDataFeed(dataFeed);

System.out.printf("Data feed Id : %s%n", createdSqlDataFeed.getId());
System.out.printf("Data feed name : %s%n", createdSqlDataFeed.getName());
System.out.printf("Is the query user is one of data feed administrator : %s%n", createdSqlDataFeed.isAdmin());
System.out.printf("Data feed created time : %s%n", createdSqlDataFeed.getCreatedTime());
System.out.printf("Data feed granularity type : %s%n",
    createdSqlDataFeed.getGranularity().getGranularityType());
System.out.printf("Data feed granularity value : %d%n",
    createdSqlDataFeed.getGranularity().getCustomGranularityValue());
System.out.println("Data feed related metric Ids:");
dataFeed.getMetricIds().forEach((metricId, metricName)
    -> System.out.printf("Metric Id : %s, Metric Name: %s%n", metricId, metricName));
System.out.printf("Data feed source type: %s%n", createdSqlDataFeed.getSourceType());

if (SQL_SERVER_DB == createdSqlDataFeed.getSourceType()) {
    System.out.printf("Data feed sql server query: %s%n",
        ((SqlServerDataFeedSource) createdSqlDataFeed.getSource()).getQuery());
}
        // Update the data feed.
        System.out.printf("Updating data feed: %s%n", dataFeed.getId());
        dataFeed = advisorAdministrationClient.updateDataFeed(dataFeed.setOptions(new DataFeedOptions()
            .setAdmins(Collections.singletonList("admin1@admin.com"))
        ));
        System.out.printf("Updated data feed admin list: %s%n",
            String.join(",", dataFeed.getOptions().getAdmins()));

        // Delete the data feed.
        System.out.printf("Deleting data feed: %s%n", dataFeed.getId());
        advisorAdministrationClient.deleteDataFeed(dataFeed.getId());
        System.out.printf("Deleted data feed%n");

        // List data feeds.
        System.out.printf("Listing data feeds%n");
        advisorAdministrationClient.listDataFeeds().forEach(dataFeedItem -> {
            System.out.printf("Data feed Id : %s%n", dataFeedItem.getId());
            System.out.printf("Data feed name : %s%n", dataFeedItem.getName());
            System.out.printf("Is the query user is one of data feed administrator : %s%n", dataFeedItem.isAdmin());
            System.out.printf("Data feed created time : %s%n", dataFeedItem.getCreatedTime());
            System.out.printf("Data feed granularity type : %s%n", dataFeedItem.getGranularity().getGranularityType());
            System.out.printf("Data feed granularity value : %d%n",
                dataFeedItem.getGranularity().getCustomGranularityValue());
            System.out.println("Data feed related metric Id's:");
            dataFeedItem.getMetricIds().forEach((metricId, metricName)
                -> System.out.printf("Metric Id : %s, Metric Name: %s%n", metricId, metricName));
            System.out.printf("Data feed source type: %s%n", dataFeedItem.getSourceType());
        });
    }
}

可使用以下命令生成应用:

gradle build

运行应用程序

使用 run 目标运行应用程序:

gradle run

参考文档 | 库源代码 | 包 (npm) | 示例

重要

Azure 建议使用最安全的可用身份验证流。 本文所述的某些身份验证流需要极高度地信任应用程序,并且附带了其他更安全的流中不存在的风险。 请仅在无法使用其他更安全的流(例如托管标识)时才使用此流。

先决条件

提示

  • 可以在 GitHub 上找到 JavaScript 指标顾问示例。
  • 指标顾问资源可能需要 10 到 30 分钟才能部署一个服务实例供你使用。 部署成功后,选择“转到资源”。 部署完成后,可以通过 Web 门户和 REST API 这两种方式开始使用指标顾问实例。
  • 可以在 Azure 门户的资源“概述”部分中找到 REST API 的 URL。 它将如下所示:
  • https://<instance-name>.cognitiveservices.azure.cn/

设置

创建新的 Node.js 应用程序

在控制台窗口(例如 cmd、PowerShell 或 Bash)中,为应用创建一个新目录并导航到该目录。

mkdir myapp && cd myapp

运行 npm init 命令以使用 package.json 文件创建一个 node 应用程序。

npm init

安装客户端库

安装 @azure/ai-metrics-advisor npm 包:

npm install @azure/ai-metrics-advisor

应用的 package.json 文件将使用依赖项进行更新。

环境变量

若要成功对异常检测器服务发出调用,需要使用以下值:

变量名称
METRICS_ADVISOR_ENDPOINT 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到此值。 终结点示例:https://YOUR_RESOURCE_NAME.cognitiveservices.azure.cn/
METRICS_ADVISOR_KEY 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到键值。 可以使用 KEY1KEY2
METRICS_ADVISOR_API_KEY 指标顾问门户检查资源时,可以在“设置”>“API 密钥”下找到密钥值。 可以使用 KEY1KEY2
SQL_CONNECTION_STRING 本快速入门要求你拥有自己的 SQL 数据库 + 连接字符串。 示例连接字符串类似于以下示例:Data Source=<Server>;Initial Catalog=<db-name>;User ID=<user-name>;Password=<password>有关构造 SQL 连接字符串的详细信息,请参阅 SQL 文档
SQL_QUERY 特定于数据集的唯一查询。

创建环境变量

为密钥和终结点创建和分配持久环境变量。

重要

如果使用 API 密钥,请将其安全地存储在某个其他位置,例如 Azure Key Vault 中。 请不要直接在代码中包含 API 密钥,并且切勿公开发布该密钥。

有关 Azure AI 服务安全性的详细信息,请参阅对 Azure AI 服务的请求进行身份验证

setx METRICS_ADVISOR_ENDPOINT "REPLACE_WITH_YOUR_ENDPOINT_HERE" 
setx METRICS_ADVISOR_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx METRICS_ADVISOR_API_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx SQL_CONNECTION_STRING "REPLACE_WITH_YOUR_UNIQUE_SQL_CONNECTION_STRING" 
setx SQL_QUERY "REPLACE_WITH_YOUR_UNIQUE_SQL_QUERY_BASED_ON_THE_UNDERLYING_STRUCTURE_OF_YOUR_DATA" 

创建应用程序

创建一个名为 index.js 的文件,并复制以下代码:

/**
 *  @summary This sample demonstrates how to get started by creating a data feed, checking ingestion status,
 * creating detection and alerting configurations, and querying for alerts and anomalies.
 */

// Load the .env file if it exists
const dotenv = require("dotenv");
dotenv.config();

const {
  MetricsAdvisorKeyCredential,
  MetricsAdvisorAdministrationClient,
  MetricsAdvisorClient
} = require("@azure/ai-metrics-advisor");

async function main() {
  // You will need to set these environment variables or edit the following values
  const endpoint = process.env["METRICS_ADVISOR_ENDPOINT"] || "<service endpoint>";
  const subscriptionKey = process.env["METRICS_ADVISOR_KEY"] || "<subscription key>";
  const apiKey = process.env["METRICS_ADVISOR_API_KEY"] || "<api key>";
  const sqlServerConnectionString =
    process.env["SQL_SERVER_CONNECTION_STRING"] ||
    "<connection string to SQL Server>";
  const sqlServerQuery =
    process.env["SQL_SERVER_QUERY"] || "<SQL Server query to retrive data>";

  const credential = new MetricsAdvisorKeyCredential(subscriptionKey, apiKey);

  const client = new MetricsAdvisorClient(endpoint, credential);
  const adminClient = new MetricsAdvisorAdministrationClient(endpoint, credential);

  const created = await createDataFeed(adminClient, sqlServerConnectionString, sqlServerQuery);
  console.log(`Data feed created: ${created.id}`);
  console.log("  metrics: ");
  console.log(created.schema.metrics);

  console.log("Waiting for a minute before checking ingestion status...");
  await delay(60 * 1000);

  try {
    await checkIngestionStatus(
      adminClient,
      created.id,
      new Date(Date.UTC(2020, 8, 1)),
      new Date(Date.UTC(2020, 8, 12))
    );

    const metricId = created.schema.metrics[0].id;
    const detectionConfig = await configureAnomalyDetectionConfiguration(adminClient, metricId);
    console.log(`Detection configuration created: ${detectionConfig.id}`);

    const hook = await createWebhookHook(adminClient);
    console.log(`Webhook hook created: ${hook.id}`);

    const alertConfig = await configureAlertConfiguration(adminClient, detectionConfig.id, [
      hook.id
    ]);
    console.log(`Alert configuration created: ${alertConfig.id}`);

    // you can use alert configuration created in above step to query the alert.
    const alerts = await queryAlerts(
      client,
      alertConfig.id,
      new Date(Date.UTC(2020, 8, 1)),
      new Date(Date.UTC(2020, 8, 12))
    );

    if (alerts.length > 1) {
      // query anomalies using an alert id.
      await queryAnomaliesByAlert(client, alerts[0]);
    } else {
      console.log("No alerts during the time period");
    }
  } finally {
    console.log(`Deleting the data feed '${created.id}`);
    await adminClient.deleteDataFeed(created.id);
  }
}

async function createDataFeed(adminClient, sqlServerConnectionString, sqlServerQuery) {
  console.log("Creating Datafeed...");
  const dataFeed = {
    name: "test_datafeed_" + new Date().getTime().toString(),
    source: {
      dataSourceType: "SqlServer",
      connectionString: sqlServerConnectionString,
      query: sqlServerQuery,
      authenticationType: "Basic"
    },
    granularity: {
      granularityType: "Daily"
    },
    schema: {
      metrics: [
        {
          name: "revenue",
          displayName: "revenue",
          description: "Metric1 description"
        },
        {
          name: "cost",
          displayName: "cost",
          description: "Metric2 description"
        }
      ],
      dimensions: [
        { name: "city", displayName: "city display" },
        { name: "category", displayName: "category display" }
      ],
      timestampColumn: undefined
    },
    ingestionSettings: {
      ingestionStartTime: new Date(Date.UTC(2020, 5, 1)),
      ingestionStartOffsetInSeconds: 0,
      dataSourceRequestConcurrency: -1,
      ingestionRetryDelayInSeconds: -1,
      stopRetryAfterInSeconds: -1
    },
    rollupSettings: {
      rollupType: "AutoRollup",
      rollupMethod: "Sum",
      rollupIdentificationValue: "__SUM__"
    },
    missingDataPointFillSettings: {
      fillType: "SmartFilling"
    },
    accessMode: "Private",
    admins: ["xyz@microsoft.com"]
  };
  const result = await adminClient.createDataFeed(dataFeed);

  return result;
}

async function checkIngestionStatus(adminClient, datafeedId, startTime, endTime) {
  // This shows how to use for-await-of syntax to list status
  console.log("Checking ingestion status...");
  const listIterator = adminClient.listDataFeedIngestionStatus(datafeedId, startTime, endTime);
  for await (const status of listIterator) {
    console.log(`  [${status.timestamp}] ${status.status} - ${status.message}`);
  }
}

async function configureAnomalyDetectionConfiguration(adminClient, metricId) {
  console.log(`Creating an anomaly detection configuration on metric '${metricId}'...`);
  const anomalyConfig = {
    name: "test_detection_configuration" + new Date().getTime().toString(),
    metricId,
    wholeSeriesDetectionCondition: {
      smartDetectionCondition: {
        sensitivity: 100,
        anomalyDetectorDirection: "Both",
        suppressCondition: {
          minNumber: 1,
          minRatio: 1
        }
      }
    },
    description: "Detection configuration description"
  };
  return await adminClient.createDetectionConfig(anomalyConfig);
}

async function createWebhookHook(adminClient) {
  console.log("Creating a webhook hook");
  const hook = {
    hookType: "Webhook",
    name: "web hook " + new Date().getTime().toString(),
    description: "description",
    hookParameter: {
      endpoint: "https://httpbin.org/post",
      username: "user",
      password: "pass"
      // certificateKey: "k",
      // certificatePassword: "kp"
    }
  };

  return await adminClient.createHook(hook);
}

async function configureAlertConfiguration(adminClient, detectionConfigId, hookIds) {
  console.log("Creating a new alerting configuration...");
  const anomalyAlert = {
    name: "test_alert_config_" + new Date().getTime().toString(),
    crossMetricsOperator: "AND",
    metricAlertConfigurations: [
      {
        detectionConfigurationId: detectionConfigId,
        alertScope: {
          scopeType: "All"
        },
        alertConditions: {
          severityCondition: {
            minAlertSeverity: "Medium",
            maxAlertSeverity: "High"
          }
        },
        snoozeCondition: {
          autoSnooze: 0,
          snoozeScope: "Metric",
          onlyForSuccessive: true
        }
      }
    ],
    hookIds,
    description: "Alerting config description"
  };
  return await adminClient.createAlertConfig(anomalyAlert);
}

async function queryAlerts(client, alertConfigId, startTime, endTime) {
  console.log(`Listing alerts for alert configuration '${alertConfigId}'`);
  // This shows how to use `for-await-of` syntax to list alerts
  console.log("  using for-await-of syntax");
  let alerts = [];
  const listIterator = client.listAlerts(alertConfigId, startTime, endTime, "AnomalyTime");
  for await (const alert of listIterator) {
    alerts.push(alert);
    console.log("    Alert");
    console.log(`      id: ${alert.id}`);
    console.log(`      timestamp: ${alert.timestamp}`);
    console.log(`      created on: ${alert.createdOn}`);
  }
  // alternatively we could list results by pages
  console.log(`  by pages`);
  const iterator = client
    .listAlerts(alertConfigId, startTime, endTime, "AnomalyTime")
    .byPage({ maxPageSize: 2 });

  let result = await iterator.next();
  while (!result.done) {
    console.log("    -- Page -- ");
    for (const item of result.value) {
      console.log(`      id: ${item.id}`);
      console.log(`      timestamp: ${item.timestamp}`);
      console.log(`      created on: ${item.createdOn}`);
    }
    result = await iterator.next();
  }

  return alerts;
}

async function queryAnomaliesByAlert(client, alert) {
  console.log(
    `Listing anomalies for alert configuration '${alert.alertConfigId}' and alert '${alert.id}'`
  );
  const listIterator = client.listAnomaliesForAlert(alert);
  for await (const anomaly of listIterator) {
    console.log(
      `  Anomaly ${anomaly.severity} ${anomaly.status} ${anomaly.seriesKey.dimension} ${anomaly.timestamp}`
    );
  }
}

async function delay(milliseconds) {
  return new Promise((resolve) => setTimeout(resolve, milliseconds));
}

main()
  .then((_) => {
    console.log("Succeeded");
  })
  .catch((err) => {
    console.log("Error occurred:");
    console.log(err);
  });

运行应用程序

在快速入门文件中使用 node 命令运行应用程序。

node index.js

参考文档 | 库源代码 | 包 (PiPy) | 示例

重要

Azure 建议使用最安全的可用身份验证流。 本文所述的某些身份验证流需要极高度地信任应用程序,并且附带了其他更安全的流中不存在的风险。 请仅在无法使用其他更安全的流(例如托管标识)时才使用此流。

先决条件

提示

  • 可以在 GitHub 上找到 Python 指标顾问示例。
  • 指标顾问资源可能需要 10 到 30 分钟才能部署一个服务实例供你使用。 部署成功后,选择“转到资源”。 部署完成后,可以通过 Web 门户和 REST API 这两种方式开始使用指标顾问实例。
  • 可以在 Azure 门户的资源“概述”部分中找到 REST API 的 URL。 它将如下所示:
  • https://<instance-name>.cognitiveservices.azure.cn/

设置

安装客户端库

安装客户端库。 可使用以下方式安装客户端库:

pip install azure-ai-metricsadvisor --pre

环境变量

若要成功对异常检测器服务发出调用,需要使用以下值:

变量名称
METRICS_ADVISOR_ENDPOINT 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到此值。 终结点示例:https://YOUR_RESOURCE_NAME.cognitiveservices.azure.cn/
METRICS_ADVISOR_KEY 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到键值。 可以使用 KEY1KEY2
METRICS_ADVISOR_API_KEY 指标顾问门户检查资源时,可以在“设置”>“API 密钥”下找到密钥值。 可以使用 KEY1KEY2
SQL_CONNECTION_STRING 本快速入门要求你拥有自己的 SQL 数据库 + 连接字符串。 示例连接字符串类似于以下示例:Data Source=<Server>;Initial Catalog=<db-name>;User ID=<user-name>;Password=<password>有关构造 SQL 连接字符串的详细信息,请参阅 SQL 文档
SQL_QUERY 特定于数据集的唯一查询。

创建环境变量

为密钥和终结点创建和分配持久环境变量。

重要

如果使用 API 密钥,请将其安全地存储在某个其他位置,例如 Azure Key Vault 中。 请不要直接在代码中包含 API 密钥,并且切勿公开发布该密钥。

有关 Azure AI 服务安全性的详细信息,请参阅对 Azure AI 服务的请求进行身份验证

setx METRICS_ADVISOR_ENDPOINT "REPLACE_WITH_YOUR_ENDPOINT_HERE" 
setx METRICS_ADVISOR_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx METRICS_ADVISOR_API_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx SQL_CONNECTION_STRING "REPLACE_WITH_YOUR_UNIQUE_SQL_CONNECTION_STRING" 
setx SQL_QUERY "REPLACE_WITH_YOUR_UNIQUE_SQL_QUERY_BASED_ON_THE_UNDERLYING_STRUCTURE_OF_YOUR_DATA" 

创建应用程序

基于以下代码创建 Python 应用程序:

"""
FILE: sample_data_feeds.py
DESCRIPTION:
    This sample demonstrates how to create, get, list, update, and delete datafeeds under your Metrics Advisor account.
USAGE:
    python sample_data_feeds.py
    Set the environment variables with your own values before running the sample:
    1) METRICS_ADVISOR_ENDPOINT - the endpoint of your Azure AI Metrics Advisor service
    2) METRICS_ADVISOR_KEY - Metrics Advisor service subscription key
    3) METRICS_ADVISOR_API_KEY - Metrics Advisor service API key
    4) SQL_CONNECTION_STRING - Used in this sample for demonstration, but you should
       add your own credentials specific to the data source type you're using
    5) SQL_QUERY - Used in this sample for demonstration, but you should
       add your own query specific to the structure of the data in your datasource.
"""

import os
import datetime


def sample_create_data_feed():
    from azure.ai.metricsadvisor import MetricsAdvisorKeyCredential, MetricsAdvisorAdministrationClient
    from azure.ai.metricsadvisor.models import (
        SqlServerDataFeedSource,
        DataFeedSchema,
        DataFeedMetric,
        DataFeedDimension,
        DataFeedRollupSettings,
        DataFeedMissingDataPointFillSettings,
    )

    service_endpoint = os.getenv("METRICS_ADVISOR_ENDPOINT")
    subscription_key = os.getenv("METRICS_ADVISOR_KEY")
    api_key = os.getenv("METRICS_ADVISOR_API_KEY")
    sql_server_connection_string = os.getenv("SQL_CONNECTION_STRING")
    query = os.getenv("SQL_QUERY")

    client = MetricsAdvisorAdministrationClient(service_endpoint,
                                  MetricsAdvisorKeyCredential(subscription_key, api_key))

    data_feed = client.create_data_feed(
        name="My data feed",
        source=SqlServerDataFeedSource(
            connection_string=sql_server_connection_string,
            query=query,
        ),
        granularity="Daily",
        schema=DataFeedSchema(
            metrics=[
                DataFeedMetric(name="cost", display_name="Cost"),
                DataFeedMetric(name="revenue", display_name="Revenue")
            ],
            dimensions=[
                DataFeedDimension(name="category", display_name="Category"),
                DataFeedDimension(name="region", display_name="region")
            ],
            timestamp_column="Timestamp"
        ),
        ingestion_settings=datetime.datetime(2019, 10, 1),
        data_feed_description="cost/revenue data feed",
        rollup_settings=DataFeedRollupSettings(
            rollup_type="AutoRollup",
            rollup_method="Sum",
            rollup_identification_value="__CUSTOM_SUM__"
        ),
        missing_data_point_fill_settings=DataFeedMissingDataPointFillSettings(
            fill_type="SmartFilling"
        ),
        access_mode="Private"
    )

    return data_feed

def sample_get_data_feed(data_feed_id):
    from azure.ai.metricsadvisor import MetricsAdvisorKeyCredential, MetricsAdvisorAdministrationClient

    service_endpoint = os.getenv("METRICS_ADVISOR_ENDPOINT")
    subscription_key = os.getenv("METRICS_ADVISOR_KEY")
    api_key = os.getenv("METRICS_ADVISOR_API_KEY")

    client = MetricsAdvisorAdministrationClient(service_endpoint,
                                  MetricsAdvisorKeyCredential(subscription_key, api_key))

    data_feed = client.get_data_feed(data_feed_id)

    print("ID: {}".format(data_feed.id))
    print("Data feed name: {}".format(data_feed.name))
    print("Created time: {}".format(data_feed.created_time))
    print("Status: {}".format(data_feed.status))
    print("Source type: {}".format(data_feed.source.data_source_type))
    print("Granularity type: {}".format(data_feed.granularity.granularity_type))
    print("Data feed metrics: {}".format([metric.name for metric in data_feed.schema.metrics]))
    print("Data feed dimensions: {}".format([dimension.name for dimension in data_feed.schema.dimensions]))
    print("Data feed timestamp column: {}".format(data_feed.schema.timestamp_column))
    print("Ingestion data starting on: {}".format(data_feed.ingestion_settings.ingestion_begin_time))
    print("Data feed description: {}".format(data_feed.data_feed_description))
    print("Data feed rollup type: {}".format(data_feed.rollup_settings.rollup_type))
    print("Data feed rollup method: {}".format(data_feed.rollup_settings.rollup_method))
    print("Data feed fill setting: {}".format(data_feed.missing_data_point_fill_settings.fill_type))
    print("Data feed access mode: {}".format(data_feed.access_mode))

def sample_list_data_feeds():
    from azure.ai.metricsadvisor import MetricsAdvisorKeyCredential, MetricsAdvisorAdministrationClient

    service_endpoint = os.getenv("METRICS_ADVISOR_ENDPOINT")
    subscription_key = os.getenv("METRICS_ADVISOR_SUBSCRIPTION_KEY")
    api_key = os.getenv("METRICS_ADVISOR_API_KEY")

    client = MetricsAdvisorAdministrationClient(service_endpoint,
                                  MetricsAdvisorKeyCredential(subscription_key, api_key))

    data_feeds = client.list_data_feeds()

    for feed in data_feeds:
        print("Data feed name: {}".format(feed.name))
        print("ID: {}".format(feed.id))
        print("Created time: {}".format(feed.created_time))
        print("Status: {}".format(feed.status))
        print("Source type: {}".format(feed.source.data_source_type))
        print("Granularity type: {}".format(feed.granularity.granularity_type))

        print("\nFeed metrics:")
        for metric in feed.schema.metrics:
            print(metric.name)

        print("\nFeed dimensions:")
        for dimension in feed.schema.dimensions:
            print(dimension.name)

def sample_update_data_feed(data_feed):
    from azure.ai.metricsadvisor import MetricsAdvisorKeyCredential, MetricsAdvisorAdministrationClient

    service_endpoint = os.getenv("METRICS_ADVISOR_ENDPOINT")
    subscription_key = os.getenv("METRICS_ADVISOR_KEY")
    api_key = os.getenv("METRICS_ADVISOR_API_KEY")

    client = MetricsAdvisorAdministrationClient(service_endpoint,
                                  MetricsAdvisorKeyCredential(subscription_key, api_key))

    # update data feed on the data feed itself or by using available keyword arguments
    data_feed.name = "updated name"
    data_feed.data_feed_description = "updated description for data feed"

    updated = client.update_data_feed(
        data_feed,
        access_mode="Public",
        fill_type="CustomValue",
        custom_fill_value=1
    )
    print("Updated name: {}".format(updated.name))
    print("Updated description: {}".format(updated.data_feed_description))
    print("Updated access mode: {}".format(updated.access_mode))
    print("Updated fill setting, value: {}, {}".format(
        updated.missing_data_point_fill_settings.fill_type,
        updated.missing_data_point_fill_settings.custom_fill_value,
    ))

def sample_delete_data_feed(data_feed_id):
    from azure.core.exceptions import ResourceNotFoundError
    from azure.ai.metricsadvisor import MetricsAdvisorKeyCredential, MetricsAdvisorAdministrationClient

    service_endpoint = os.getenv("METRICS_ADVISOR_ENDPOINT")
    subscription_key = os.getenv("METRICS_ADVISOR_KEY")
    api_key = os.getenv("METRICS_ADVISOR_API_KEY")

    client = MetricsAdvisorAdministrationClient(service_endpoint,
                                  MetricsAdvisorKeyCredential(subscription_key, api_key))

    client.delete_data_feed(data_feed_id)

    try:
        client.get_data_feed(data_feed_id)
    except ResourceNotFoundError:
        print("Data feed successfully deleted.")

if __name__ == '__main__':
    print("---Creating data feed...")
    data_feed = sample_create_data_feed()
    print("Data feed successfully created...")
    print("\n---Get a data feed...")
    sample_get_data_feed(data_feed.id)
    print("\n---List data feeds...")
    sample_list_data_feeds()
    print("\n---Update a data feed...")
    sample_update_data_feed(data_feed)
    print("\n---Delete a data feed...")
    sample_delete_data_feed(data_feed.id)

运行应用程序

在快速入门文件中使用 python 命令运行应用程序。

python quickstart-file.py

先决条件

重要

Azure 建议使用最安全的可用身份验证流。 本文所述的某些身份验证流需要极高度地信任应用程序,并且附带了其他更安全的流中不存在的风险。 请仅在无法使用其他更安全的流(例如托管标识)时才使用此流。

设置

本快速入门的示例代码将演示如何使用 Python 调用 REST API。 有关特定的 REST API 调用,请参阅 GitHub 示例

提示

  • 指标顾问资源可能需要 10 到 30 分钟才能部署一个服务实例供你使用。 部署成功后,选择“转到资源”。 部署完成后,可以通过 Web 门户和 REST API 这两种方式开始使用指标顾问实例。
  • 可以在 Azure 门户的资源“概述”部分中找到 REST API 的 URL。 它将如下所示:
  • https://<instance-name>.cognitiveservices.azure.cn/

环境变量

若要成功对异常检测器服务发出调用,需要使用以下值:

变量名称
METRICS_ADVISOR_ENDPOINT 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到此值。 终结点示例:https://YOUR_RESOURCE_NAME.cognitiveservices.azure.cn/
METRICS_ADVISOR_KEY 从 Azure 门户检查资源时,可在“密钥和终结点”部分中找到键值。 可以使用 KEY1KEY2
METRICS_ADVISOR_API_KEY 指标顾问门户检查资源时,可以在“设置”>“API 密钥”下找到密钥值。 可以使用 KEY1KEY2
SQL_CONNECTION_STRING 本快速入门要求你拥有自己的 SQL 数据库 + 连接字符串。 示例连接字符串类似于以下示例:Data Source=<Server>;Initial Catalog=<db-name>;User ID=<user-name>;Password=<password>有关构造 SQL 连接字符串的详细信息,请参阅 SQL 文档
SQL_QUERY 特定于数据集的唯一查询。

创建环境变量

为密钥和终结点创建和分配持久环境变量。

重要

如果使用 API 密钥,请将其安全地存储在某个其他位置,例如 Azure Key Vault 中。 请不要直接在代码中包含 API 密钥,并且切勿公开发布该密钥。

有关 Azure AI 服务安全性的详细信息,请参阅对 Azure AI 服务的请求进行身份验证

setx METRICS_ADVISOR_ENDPOINT "REPLACE_WITH_YOUR_ENDPOINT_HERE" 
setx METRICS_ADVISOR_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx METRICS_ADVISOR_API_KEY "REPLACE_WITH_YOUR_KEY_VALUE_HERE" 
setx SQL_CONNECTION_STRING "REPLACE_WITH_YOUR_UNIQUE_SQL_CONNECTION_STRING" 
setx SQL_QUERY "REPLACE_WITH_YOUR_UNIQUE_SQL_QUERY_BASED_ON_THE_UNDERLYING_STRUCTURE_OF_YOUR_DATA" 

创建应用程序

import requests
import json
import time


def add_data_feed(endpoint, subscription_key, api_key):
    url = endpoint + '/dataFeeds'
    data_feed_body = {
        "dataSourceType": "SqlServer",
        "dataFeedName": "test_data_feed_00000001",
        "dataFeedDescription": "",
        "dataSourceParameter": {
            "connectionString": os.environ['SQL_CONNECTION_STRING'],
            "query": os.environ['SQL_QUERY']
        },
        "granularityName": "Daily",
        "granularityAmount": 0,
        "metrics": [
            {
                "metricName": "revenue",
                "metricDisplayName": "revenue",
                "metricDescription": ""
            },
            {
                "metricName": "cost",
                "metricDisplayName": "cost",
                "metricDescription": ""
            }
        ],
        "dimension": [
            {
                "dimensionName": "city",
                "dimensionDisplayName": "city"
            },
            {
                "dimensionName": "category",
                "dimensionDisplayName": "category"
            }
        ],
        "timestampColumn": "timestamp",
        "dataStartFrom": "2020-06-01T00:00:00.000Z",
        "startOffsetInSeconds": 0,
        "maxConcurrency": -1,
        "minRetryIntervalInSeconds": -1,
        "stopRetryAfterInSeconds": -1,
        "allUpIdentification": "__SUM__",
        "needRollup": "AlreadyRollup",
        "fillMissingPointType": "SmartFilling",
        "fillMissingPointValue": 0,
        "viewMode": "Private",
        "admins": [
            "admin@contoso.com"
        ],
        "viewers": [
        ],
        "actionLinkTemplate": ""
    }
    res = requests.post(url, data=json.dumps(data_feed_body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 201:
        raise RuntimeError("add_data_feed failed " + res.text)
    else:
        print("add_data_feed success " + res.text)
    return res.headers['Location']


def check_ingestion_latest_status(endpoint, subscription_key, api_key, datafeed_id):
    url = endpoint + '/dataFeeds/{}/ingestionProgress'.format(datafeed_id)
    res = requests.get(url, headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                     'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("check_ingestion_latest_status failed " + res.text)
    else:
        print("check_ingestion_latest_status success " + res.text)


def check_ingestion_detail_status(endpoint, subscription_key, api_key, datafeed_id, start_time, end_time):
    url = endpoint + '/dataFeeds/{}/ingestionStatus/query'.format(datafeed_id)
    ingestion_detail_status_body = {
      "startTime": start_time,
      "endTime": end_time
    }
    res = requests.post(url, data=json.dumps(ingestion_detail_status_body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("check_ingestion_detail_status failed " + res.text)
    else:
        print("check_ingestion_detail_status success " + res.text)


def create_detection_config(endpoint, subscription_key, api_key, metric_id):
    url = endpoint + '/enrichment/anomalyDetection/configurations'
    detection_config_body = {
        "name": "test_detection_config0000000001",
        "description": "string",
        "metricId": metric_id,
        "wholeMetricConfiguration": {
            "smartDetectionCondition": {
                "sensitivity": 100,
                "anomalyDetectorDirection": "Both",
                "suppressCondition": {
                    "minNumber": 1,
                    "minRatio": 1
                }
            }
        },
        "dimensionGroupOverrideConfigurations": [
        ],
        "seriesOverrideConfigurations": [
        ]
    }
    res = requests.post(url, data=json.dumps(detection_config_body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 201:
        raise RuntimeError("create_detection_config failed " + res.text)
    else:

        print("create_detection_config success " + res.text)
    return res.headers['Location']


def create_web_hook(endpoint, subscription_key, api_key):
    url = endpoint + '/hooks'
    web_hook_body = {
        "hookType": "Webhook",
        "hookName": "test_web_hook000001",
        "description": "",
        "externalLink": "",
        "hookParameter": {
            "endpoint": "https://www.contoso.com",
            "username": "",
            "password": ""
        }
    }
    res = requests.post(url, data=json.dumps(web_hook_body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 201:
        raise RuntimeError("create_web_hook failed " + res.text)
    else:
        print("create_web_hook success " + res.text)
    return res.headers['Location']


def create_alert_config(endpoint, subscription_key, api_key, anomaly_detection_configuration_id, hook_id):
    url = endpoint + '/alert/anomaly/configurations'
    web_hook_body = {
        "name": "test_alert_config00000001",
        "description": "",
        "crossMetricsOperator": "AND",
        "hookIds": [
           hook_id
        ],
        "metricAlertingConfigurations": [
            {
                "anomalyDetectionConfigurationId": anomaly_detection_configuration_id,
                "anomalyScopeType": "All",
                "severityFilter": {
                    "minAlertSeverity": "Low",
                    "maxAlertSeverity": "High"
                },
                "snoozeFilter": {
                    "autoSnooze": 0,
                    "snoozeScope": "Metric",
                    "onlyForSuccessive": True
                },
            }
        ]
    }
    res = requests.post(url, data=json.dumps(web_hook_body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 201:
        raise RuntimeError("create_alert_config failed " + res.text)
    else:
        print("create_alert_config success " + res.text)
    return res.headers['Location']


def query_alert_by_alert_config(endpoint, subscription_key, api_key, alert_config_id, start_time, end_time):
    url = endpoint + '/alert/anomaly/configurations/{}/alerts/query'.format(alert_config_id)
    alerts_body = {
        "startTime": start_time,
        "endTime": end_time,
        "timeMode": "AnomalyTime"
    }
    res = requests.post(url, data=json.dumps(alerts_body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("query_alert_by_alert_config failed " + res.text)
    else:
        print("query_alert_by_alert_config success " + res.text)
    return [item['alertId'] for item in json.loads(res.content)['value']]


def query_anomaly_by_alert(endpoint, subscription_key, api_key, alert_config_id, alert_id):
    url = endpoint + '/alert/anomaly/configurations/{}/alerts/{}/anomalies'.format(alert_config_id, alert_id)
    res = requests.get(url,
                       headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("query_anomaly_by_alert failed " + res.text)
    else:
        print("query_anomaly_by_alert success " + res.text)
    return json.loads(res.content)


def query_incident_by_alert(endpoint, subscription_key, api_key, alert_config_id, alert_id):
    url = endpoint + '/alert/anomaly/configurations/{}/alerts/{}/incidents'.format(alert_config_id, alert_id)
    res = requests.get(url,
                       headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("query_incident_by_alert failed " + res.text)
    else:
        print("query_incident_by_alert success " + res.text)
    return json.loads(res.content)


def query_root_cause_by_incident(endpoint, subscription_key, api_key, detection_config_id, incident_id):
    url = endpoint + '/enrichment/anomalyDetection/configurations/{}/incidents/{}/rootCause'.format(detection_config_id, incident_id)
    res = requests.get(url,
                       headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("query_root_cause_by_incident failed " + res.text)
    else:
        print("query_root_cause_by_incident success " + res.text)
    return json.loads(res.content)


def query_anomaly_by_detection_config(endpoint, subscription_key, api_key, detection_config_id, start_time, end_time):
    url = endpoint + '/enrichment/anomalyDetection/configurations/{}/anomalies/query'.format(detection_config_id)
    body = {
        "startTime": start_time,
        "endTime": end_time,
        "filter": {
            "dimensionFilter": [
            ],
            "severityFilter": {
              "min": "Low",
              "max": "High"
            }
          }
    }
    res = requests.post(url, data=json.dumps(body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("query_anomaly_by_detection_config failed " + res.text)
    else:
        print("query_anomaly_by_detection_config success " + res.text)
    return json.loads(res.content)


def query_incident_by_detection_config(endpoint, subscription_key, api_key, detection_config_id, start_time, end_time):
    url = endpoint + '/enrichment/anomalyDetection/configurations/{}/incidents/query'.format(detection_config_id)
    body = {
        "startTime": start_time,
        "endTime": end_time,
        "filter": {
            "dimensionFilter": [
            ],
        }
    }
    res = requests.post(url, data=json.dumps(body),
                        headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                 'x-api-key': api_key})
    if res.status_code != 200:
        raise RuntimeError("query_incident_by_detection_config failed " + res.text)
    else:
        print("query_incident_by_detection_config success " + res.text)
    return json.loads(res.content)


if __name__ == '__main__':
    # Example endpoint: https://[placeholder].cognitiveservices.azure.cn/metricsadvisor/v1.0
    endpoint = os.environ['METRICS_ADVISOR_ENDPOINT'] + "metricsadvisor/v1.0"
    subscription_key = os.environ['METRICS_ADVISOR_KEY']
    api_key = os.environ['METRICS_ADVISOR_API_KEY']

    '''
    First part
    1.onboard datafeed
    2.check datafeed latest status
    3.check datafeed status details
    4.create detection config
    5.create webhook
    6.create alert config
    '''
    datafeed_resource_url = add_data_feed(endpoint, subscription_key, api_key)
    print(datafeed_resource_url)

    # datafeed_id and metrics_id can get from datafeed_resource_url
    datafeed_info = json.loads(requests.get(url=datafeed_resource_url,
                                            headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                                     'x-api-key': api_key}).content)
    print(datafeed_info)
    datafeed_id = datafeed_info['dataFeedId']
    metrics_id = []
    for metrics in datafeed_info['metrics']:
        metrics_id.append(metrics['metricId'])
    time.sleep(60)

    check_ingestion_latest_status(endpoint, subscription_key, api_key, datafeed_id)

    check_ingestion_detail_status(endpoint, subscription_key, api_key, datafeed_id,
                                  "2020-06-01T00:00:00Z", "2020-07-01T00:00:00Z")

    detection_config_resource_url = create_detection_config(endpoint, subscription_key, api_key, metrics_id[0])
    print(detection_config_resource_url)

    # anomaly_detection_configuration_id can get from detection_config_resource_url
    detection_config = json.loads(requests.get(url=detection_config_resource_url,
                                               headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                                        'x-api-key': api_key}).content)
    print(detection_config)
    anomaly_detection_configuration_id = detection_config['anomalyDetectionConfigurationId']

    webhook_resource_url = create_web_hook(endpoint, subscription_key, api_key)
    print(webhook_resource_url)

    # hook_id can get from webhook_resource_url
    webhook = json.loads(requests.get(url=webhook_resource_url,
                                      headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                               'x-api-key': api_key}).content)
    print(webhook)
    hook_id = webhook['hookId']

    alert_config_resource_url = create_alert_config(endpoint, subscription_key, api_key,
                                                    anomaly_detection_configuration_id, hook_id)

    # anomaly_alerting_configuration_id can get from alert_config_resource_url
    alert_config = json.loads(requests.get(url=alert_config_resource_url,
                                           headers={'Ocp-Apim-Subscription-Key': subscription_key,
                                                    'x-api-key': api_key}).content)
    print(alert_config)
    anomaly_alerting_configuration_id = alert_config['anomalyAlertingConfigurationId']

清理资源

如果想要清理并移除 Azure AI 服务订阅,可以删除资源或资源组。 删除资源组同时也会删除与之相关联的任何其他资源。

后续步骤