什么是指标顾问(预览版)?What is Metrics Advisor (preview)?

指标顾问是 Azure 认知服务的一部分,使用 AI 对时序数据执行数据监视和异常情况检测。Metrics Advisor is a part of Azure Cognitive Services that uses AI perform data monitoring and anomaly detection in time series data. 该服务自动执行将模型应用于数据的过程,并提供了一组 API 和基于 Web 的工作区,用于数据引入、异常检测和诊断,无需了解机器学习即可使用它们。The service automates the process of applying models to your data, and provides a set of APIs web-based workspace for data ingestion, anomaly detection, and diagnostics - without needing to know machine learning. 使用指标顾问执行以下操作:Use Metrics Advisor to:

  • 分析来自多个数据源的多维数据Analyze multi-dimensional data from multiple data sources
  • 识别并关联异常Identify and correlate anomalies
  • 配置并微调用于数据的异常情况检测模型Configure and fine-tune the anomaly detection model used on your data
  • 诊断异常,并帮助进行根本原因分析。Diagnose anomalies and help with root cause analysis.

指标顾问概述

连接到各种数据源Connect to a variety of data sources

指标顾问可以连接到多个数据存储,并从这些数据存储中引入多维指标数据,包括:SQL Server、Azure Blob 存储、MongoDB 等。Metrics Advisor can connect to, and ingest multi-dimensional metric data from many data stores, including: SQL Server, Azure Blob Storage, MongoDB and more.

易于使用且可自定义的异常情况检测Easy-to-use and customizable anomaly detection

  • 指标顾问会自动为你的数据选择最佳模型,而无需了解任何机器学习。Metrics Advisor automatically selects the best model for your data, without needing to know any machine learning.
  • 多维指标中自动监视每个时序。Automatically monitor every time series within multi-dimensional metrics.
  • 使用参数优化交互式反馈自定义应用于数据的模型以及将来的异常情况检测结果。Use parameter tuning and interactive feedback to customize the model applied on your data, and future anomaly detection results.

通过多个通道发送实时警报Real time alerts through multiple channels

每当检测到异常时,指标顾问可以使用挂钩(例如:电子邮件挂钩、Webhook 和 Azure DevOps 挂钩)通过多个通道发送实时警报Whenever anomalies are detected, Metrics Advisor is able to send real time alerts through multiple channels using hooks, such as: email hooks, web hooks and Azure DevOps hooks. 通过灵活的警报规则可自定义要发送的警报和位置。Flexible alert rules let you customize which alerts are sent, and where.

通过分析异常生成智能诊断见解Smart diagnostic insights by analyzing anomalies

分析在多维指标上检测到的异常,并生成智能诊断见解,包括最可能的根本原因、诊断树、指标钻取等。Analyze anomalies detected on multi-dimensional metrics, and generate smart diagnostic insights including most the most likely root cause, diagnostic trees, metric drilling, and more. 通过配置指标图,可以启用跨指标分析来帮助你可视化事件。By configuring Metrics graph, cross metrics analysis can enabled to help you visualize incidents.

典型工作流Typical workflow

此工作流非常简单:载入数据后,可以微调异常检测,并创建适合场景的配置。The workflow is simple: after onboarding your data, you can fine-tune the anomaly detection, and create configurations to fit your scenario.

  1. 为指标顾问创建 Azure 资源Create an Azure resource for Metrics Advisor.
  2. 使用 Web 门户生成首个监视器。Build your first monitor using the web portal.
    1. 载入数据Onboard your data
    2. 微调异常情况检测Fine-tune anomaly detection
    3. 订阅警报Subscribe to alerts
    4. 查看诊断见解View diagnostic insights
  3. 使用 REST API 自定义实例。Use the REST API to customize your instance.

后续步骤Next steps