Azure Monitor
Azure Monitor is Azure's unified platform for collecting, analyzing, and acting on metrics, logs, and traces from Azure resources, applications, and infrastructure. It underpins Log Analytics for querying with Kusto Query Language (KQL), Application Insights for application performance monitoring, and alert rules that trigger notifications or automated actions when defined conditions are met.
Azure Monitor
Azure Monitor is the umbrella service collecting metrics and logs from every Azure resource, feeding dashboards, alerts, and Log Analytics queries.
Why It Matters in Production
Hotstar's on-call team relies on Azure Monitor dashboards showing real-time CPU, memory, and request latency across every encoding VM during live IPL matches, catching degradation before viewers notice buffering.
az monitor metrics list --resource
--metric "Percentage CPU" --interval PT1M
RememberPlatform metrics are collected automatically at no extra config, but application-level telemetry requires installing the Application Insights SDK or agent.
Frequently Asked Questions
What's the difference between Azure Monitor, Log Analytics, and Application Insights?
Azure Monitor is the umbrella platform; Log Analytics and Application Insights are components under it. Log Analytics stores and queries logs and metrics via Kusto Query Language (KQL) across infrastructure and platform resources. Application Insights is the APM layer specifically for application code — request traces, dependency calls, exception tracking. Both ultimately write into the same underlying Log Analytics workspace, so a KQL query can join infrastructure logs with application telemetry in one place.
What's a gotcha with Azure Monitor alert rules in production?
Alert rules are billed per signal evaluated and per notification sent, and teams that alert on every metric at a 1-minute frequency across dozens of resources can rack up unexpected cost. A more common operational mistake is setting an alert's severity and action group but not testing the action group itself — if the webhook or email integration is misconfigured, the alert fires silently in the portal with nobody ever notified, exactly like a dead-letter queue nobody checks.