Monitoring, Resilience, and Operational Lifecycle
Observability
Azure Monitor is the central observability platform. Metrics are numeric time-series data; logs are stored in Log Analytics workspaces. Application Insights monitors application performance and dependencies. Diagnostic settings route resource logs and metrics to destinations. Activity Log records control-plane operations. Alerts use action groups to notify or automate responses.
Resilience and recovery
Availability zones and region pairs improve resilience. Azure Backup provides point-in-time recovery; Azure Site Recovery provides disaster recovery replication and orchestration. Health probes and autoscale keep applications available. SLAs are service commitments; composite SLA is the product of dependent service SLAs. RTO and RPO drive backup and replication design.
Operational lifecycle
Advisor, Service Health, and update management support operations. Infrastructure as code, blue-green deployment, and canary releases reduce change risk. Test failover regularly; an untested backup is not a recovery plan.
Traps
Candidates often confuse metrics with logs, Activity Log with resource diagnostic logs, and Backup with Site Recovery. Backup protects data; Site Recovery orchestrates failover. Composite SLA math is frequently tested: two dependent services at 99.9% yield about 99.8%, not 99.9%. Also remember that budgets and alerts do not automatically remediate incidents unless actions are configured.
Worked example
An app depends on an App Service at 99.95% and a SQL Database at 99.99%. The composite SLA is 0.9995 × 0.9999 ≈ 99.94%. Adding a redundant path can improve availability, but only if the design avoids shared dependencies.
Rule of thumb
Define RTO and RPO first, then choose backup, replication, and failover mechanisms, and test them on a schedule.