Ops Capacity Report

2. Measure Current Usage

For each service, capture current utilization:

• CPU: average and peak utilization percentage
• Memory: average and peak usage vs allocated
• Storage: current usage vs quota, growth rate per month
• Network: bandwidth usage, request volume per second
• Connections: database connections used vs pool size
• Response time: p50, p95, p99 latencies

Flag any metric above 70% utilization as approaching capacity.

3. Analyze Growth Trends

Using the last 3-6 months of data:

• Calculate month-over-month growth rate for each metric
• Identify seasonal patterns (if any)
• Project when each resource will hit capacity at current growth

Time to capacity = (Limit - Current) / Monthly Growth Rate

4. Identify Bottlenecks

Determine which resource will hit its limit first:

• Database connections: often the first bottleneck
• Memory: especially for Node.js services
• Storage: for databases with growing datasets
• API rate limits: for third-party service dependencies
• Cold starts: for serverless functions at scale

Rank bottlenecks by time-to-capacity (shortest first).

5. Model Scaling Scenarios

Create projections for three scenarios:

6. Estimate Costs

For each scenario and time horizon, calculate infrastructure costs:

• Current monthly cost
• Projected cost at each milestone
• Cost per user/request at current and projected scale
• Cost optimization opportunities (reserved instances, caching, CDN)

7. Recommend Scaling Actions

Produce a prioritized action plan:

Include both reactive (fix when it breaks) and proactive (fix before it breaks) recommendations.

Output

• Infrastructure inventory table
• Current utilization metrics with capacity percentages
• Growth projections for 3/6/12 months
• Bottleneck analysis ranked by time-to-capacity
• Cost projections for three growth scenarios
• Prioritized scaling action plan with triggers and costs

Review Checklist

• All services inventoried with current allocation
• Usage metrics captured for CPU, memory, storage, network
• Growth trends calculated from historical data
• Bottlenecks identified and ranked by urgency
• Three scaling scenarios modeled (conservative, expected, aggressive)
• Cost projections include optimization opportunities
• Action plan has clear triggers and lead times