Skip to main content

Cloud Monitoring

Beak NOC Cloud Monitoring integrates with major public cloud providers to bring cloud resource health into the same operational view as your on-premises network. Instead of context-switching between cloud consoles and your NOC, cloud resource status, instance health, and service availability are visible directly within Beak NOC.

Supported Cloud Providers​

ProviderIntegration Method
AWSIAM Role or Access Key / Secret Key with read-only permissions
AzureService Principal with Reader role on the target subscription
GCPService Account with Monitoring Viewer role

Adding a Cloud Provider​

Click Add Provider to connect a cloud account. The setup wizard steps through:

  1. Select Provider — Choose AWS, Azure, or GCP
  2. Credentials — Enter API credentials or configure role-based access
  3. Regions — Select which cloud regions to monitor (all regions or a subset)
  4. Resources — Choose which resource types to import (compute instances, load balancers, databases, Kubernetes clusters, etc.)
  5. Polling Interval — Set how frequently Beak NOC queries the cloud APIs for updated metrics

Once configured, cloud resources appear in the monitoring inventory alongside on-premises devices.

Cloud Resource Monitoring​

Compute Instances​

For each monitored VM or EC2/Azure VM/GCE instance:

  • State — Running, Stopped, Terminated
  • CPU Utilization — Average and peak CPU % over the monitoring period
  • Network In/Out — Bytes transferred per the cloud provider's metrics
  • Status Checks — Provider-level health checks (system status and instance status)

Cloud Services​

Beyond compute, Beak NOC tracks the health of managed cloud services:

Service TypeExamples
Load BalancersAWS ALB/NLB, Azure Application Gateway
DatabasesAWS RDS, Azure SQL, Cloud SQL
StorageS3 bucket availability, Azure Blob errors
KubernetesNode readiness, pod failure counts

Cloud Alerts​

Cloud-native alerts (AWS CloudWatch Alarms, Azure Monitor Alerts) are ingested and unified into the Beak NOC Alerts feed so all alerts — cloud and on-premises — appear in a single queue.

Unified Network + Cloud View​

When a network issue affects cloud connectivity (for example, an SD-WAN tunnel to a cloud region goes down), Beak NOC correlates:

  • The WAN/SD-WAN alert on the network side
  • The increased error rate on the cloud load balancer
  • The application availability drop in App Monitor

This unified view eliminates the need to manually correlate cloud console events with network events.

Best Practices​

  • Use read-only credentials — Beak NOC only requires read access to cloud APIs. Never grant write permissions to the monitoring integration.
  • Monitor all regions where you have active workloads — Cloud providers isolate region-specific issues; a problem in us-east-1 may not be visible from metrics in eu-west-1.
  • Set instance-specific thresholds — A database server running at 80% CPU is normal during a batch job; the same utilization on an application server may indicate a problem.
  • Review cloud alerts daily — Cloud-native alerts often have different definitions than network alerts. Review the alert taxonomy when first integrating a provider to avoid alert flood.
  • App Monitor — Application-level monitoring for cloud-hosted applications
  • SD-WAN — SD-WAN tunnel health for cloud on-ramp connections
  • Alerts — Unified alert feed including cloud-native alerts
  • Reports — Cloud resource availability included in monthly summary reports