Usage Analysis

The Usage Analysis page provides deep insights into resource utilization, helping you identify underutilized resources and optimization opportunities.

Accessing Usage Analysis

Navigate to FinOps > Usage Analysis in the sidebar, or visit /finops/usage.

Overview Tab

Efficiency Metrics

Key metrics at a glance:

Resource Distribution

Charts showing:

• Resource count by type (EC2, RDS, EBS, etc.)
• Resource count by utilization tier (High, Medium, Low, Idle)
• Cost distribution by resource state

Quick Insights

Automated insights including:

• Idle resources detected in the last 7 days
• Resources with consistently low CPU utilization
• Unattached volumes and unused elastic IPs

Resources Tab

Resource Inventory

Complete inventory of monitored cloud resources:

Filters

Filter resources by:

• Cloud connection
• Resource type
• Region
• State (running, stopped, terminated)
• Tags

Resource Details

Click on a resource to see:

• Full resource metadata
• Utilization metrics over time
• Cost breakdown
• Associated costs from other services (e.g., EBS volumes for EC2)

Utilization Tab

Metrics Collected

Realm9 collects the following metrics from CloudWatch (AWS) and Azure Monitor:

Compute Resources (EC2/VMs):

• CPU Utilization (%)
• Network In/Out (bytes)
• Disk Read/Write (bytes)
• Disk Read/Write Operations

Database Resources (RDS):

• CPU Utilization (%)
• Database Connections
• Read/Write IOPS
• Free Storage Space

Storage (EBS/Disks):

• Volume Read/Write Bytes
• Volume Read/Write Ops
• Throughput
• Queue Length

Utilization Charts

Interactive charts showing:

• CPU Utilization Heatmap: Time vs. resource matrix
• Network Traffic: Aggregated in/out over time
• Disk Activity: Read/write patterns

Utilization Tiers

Resources are classified into tiers:

Time Range

Analyze utilization over:

• Last 24 hours
• Last 7 days
• Last 30 days
• Custom range

Rightsizing Tab

Rightsizing Recommendations

Automated recommendations based on utilization data:

Recommendation Details

Each recommendation includes:

• Current State: Instance type, utilization metrics
• Recommended State: Target instance type
• Estimated Savings: Monthly cost reduction
• Risk Level: Impact assessment (Low/Medium/High)
• Evidence: Charts showing utilization data

Implementation

For each recommendation:

1. Review the evidence and risk level
2. Click Implement to create a change request
3. Optionally, create a Terraform plan for IaC environments
4. Track implementation status

Savings Tracking

After implementing recommendations:

• Track actual vs. estimated savings
• Monitor utilization of rightsized resources
• Get alerts if utilization changes significantly

Metrics Collection

Data Sources

Data Retention

• Raw metrics: 7 days
• Hourly aggregates: 90 days
• Daily aggregates: 365 days

Metric Lag

Utilization data may lag behind real-time by:

• 1-2 hours for CloudWatch metrics
• 5-10 minutes for Azure Monitor

Best Practices

Regular Review Cadence

Weekly:

• Check for new idle resources
• Review high utilization resources for capacity

Monthly:

• Process rightsizing recommendations
• Update utilization thresholds if needed
• Track savings from implemented changes

Rightsizing Workflow

1. Start with idle resources (lowest risk)
2. Move to low utilization (downsize opportunities)
3. Address high utilization (upsize to prevent issues)
4. Consider reserved instances for stable workloads

Accuracy Tips

• Allow 7+ days of utilization data before making decisions
• Account for batch jobs and scheduled tasks
• Consider business cycles (end of month, quarterly)

API Access

# Get resource inventory
curl -X GET https://realm9.app/api/finops/resources \
  -H "Authorization: Bearer $TOKEN"

# Get utilization metrics
curl -X GET https://realm9.app/api/finops/utilization?resource_id=i-xxx \
  -H "Authorization: Bearer $TOKEN"

# Get rightsizing recommendations
curl -X GET https://realm9.app/api/finops/rightsizing \
  -H "Authorization: Bearer $TOKEN"

Related Documentation

• FinOps Overview
• Cost Explorer
• Anomaly Detection