Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Azure Monitor is Microsoft's unified observability service for collecting, analyzing, and acting on telemetry from cloud and hybrid environments. It enables you to understand the health, performance, and reliability of your Azure applications and infrastructure resources by bringing together metrics, logs, traces, and events into a single observability experience.
In addition to supporting Azure Monitor features, the Azure Monitor data platform also supports other services such as Defender for Cloud and Microsoft Sentinel. This simplifies your management infrastructure allowing you to leverage your investment in a common set of query languages and analysis tools.
Monitor your cloud resources
Azure Monitor collects logs and metrics from all the resources across the different Azure services that you use, including audit logs from Microsoft Entra ID. This monitoring data is integrated into the Azure portal experience for each service. View a summary of the health and performance of your resources and then drill down into more detailed monitoring data and analysis tools to investigate and troubleshoot issues. Some services also provide dedicated monitoring experiences called Azure Monitor insights that deliver deep visibility into the performance and operation of your cloud resources.
Monitor your infrastructure
Azure Monitor monitors the health and performance of your virtual machines and Kubernetes clusters that make up the infrastructure of many environments. Collect event logs in addition to OpenTelemetry and Prometheus metrics and drill down on different components to view their configuration, operation, and performance details. Azure Network Watcher provides a suite of tools that use the Azure Monitor data platform to monitor the network health of your infrastructure resources.
See Monitor Azure Virtual Machines, Monitor Kubernetes, Network Watcher.
Monitor your applications
Application Insights is an OpenTelemetry feature of Azure Monitor that offers application performance monitoring (APM) for live web applications. Integrating with OpenTelemetry (OTel) provides a vendor-neutral approach to collecting and analyzing telemetry data, enabling comprehensive observability of your applications.
See Application Insights.
Monitor your AI agents
Azure Monitor's Application Insights provides a unified experience for monitoring AI agents across Microsoft Foundry, Copilot Studio, and third‑party agent frameworks. It consolidates telemetry and diagnostics, enabling developers to track agent performance, troubleshoot errors, and optimize agent behavior. Through its integration with Microsoft Foundry, Application Insights delivers real‑time observability for operators of generative AI workloads, with built‑in dashboards that surface key operational metrics such as token consumption, latency, error rates, and quality scores.
See Monitor AI agents and Observability in generative AI.
Agentic workflows
Azure Monitor combines a broad telemetry platform with AI-assisted workflows for exploration, investigation, and response. You can use the Azure Copilot Observability Agent to ask questions in natural language, investigate issues across application and infrastructure signals, and preserve findings in Azure Monitor issues. The Observability Agent helps turn raw telemetry and alert noise into explained findings with supporting evidence and recommended next steps. Run the agent on-demand when you start a chat or investigation, or have it run autonomously to act on Azure Monitor alerts. Humans remain in control of decisions and any changes to your environment.
See Azure Copilot Observability Agent.
Azure Monitor data platform
Azure Monitor has a centralized data platform to support collection of telemetry from a variety of sources and the different features for analysis and troubleshooting. The platform uses two workspace types, each optimized for different data:
- Log Analytics workspaces collect log and trace data, which can be analyzed with Kusto Query Language (KQL).
- Azure Monitor workspaces collect Prometheus and OpenTelemetry metrics, which can be analyzed using Prometheus Query Language (PromQL).
Despite the similar names, these are separate resource types with different data stores and query languages.
See Azure Monitor data platform.
Analyze and visualize data
Azure Monitor provides multiple options to visualize your monitoring data and perform ad-hoc analysis, in addition to integrating monitoring data into the Azure portal experience for different services. Use prebuilt workbooks and Grafana dashboards in the Azure portal or create your own custom dashboards. For ad-hoc analysis, use metrics explorer to analyze metrics using PromQL and Log Analytics to analyze log and trace data using KQL.
See Visualize data, metrics explorer, Log Analytics.
Troubleshooting and diagnostics
Azure Monitor provides interactive analysis tools such as metrics explorer and Log Analytics to investigate failures, correlate events, and identify performance bottlenecks across applications and resources.
See Azure Copilot Observability Agent.
Respond to issues
Alerts in Azure Monitor proactively notify you when issues are identified in collected data so that you can respond to them before they affect your customers. Azure Monitor includes AIOps capabilities that use machine learning to reduce alert noise. Features include dynamic thresholds that adapt to metric patterns and smart detection that identifies application anomalies.
The Azure Copilot Observability Agent can also act on alerts through autonomous operations. When enabled, the agent analyzes alerts in the background, correlates related alerts when they likely represent the same incident, creates Azure Monitor issues automatically, and runs deep investigations on agent-created issues. This automatic triage reduces alert noise by turning streams of individual alerts into higher-signal issues with context, findings, and recommended next steps for your on-call team to review. Teams can then act on issues through Azure Monitor action groups, including no-code workflows with Logic Apps and low-code runbooks with Azure Automation.
For more information, see Azure Monitor alerts, Smart detection, Azure Monitor issues, and Autonomous operations in the Azure Copilot Observability Agent.
Autoscale
Autoscale is an Azure Monitor feature that automatically adds and removes resources according to the load on your application. You can configure autoscale rules based on metrics, schedules, or a combination of both.
See Autoscale.
Hybrid environments
Azure Monitor isn't only designed to monitor Azure environments but can act as a centralized monitoring solution for your hybrid environments. Use Azure Arc to connect your resources in other clouds and on-premises to monitor them alongside your Azure resources. For particularly large data volumes and to support environments with intermittent connectivity, deploy the Azure Monitor pipeline which extends the data collection capabilities of Azure Monitor into your own data center and other cloud providers.
See Multicloud monitoring and Azure Monitor pipeline.
Onboarding
Onboard resources to Azure Monitor using a variety of methods including the Azure portal, command line, or infrastructure as code (IaC) templates to deploy at scale. Use the Azure Monitor Agent (AMA) installed on your virtual machines and servers to collect logs and metrics according to your data collection rules.
See Data sources and data collection methods and Resource Manager template samples.