Capacity Planning Worksheet

Forecast resource needs from growth and load tests.

Added October 15, 2026. Written with AI assistance and checked for clarity and placeholders; not reviewed by a lawyer, accountant, clinician or other professional. How these prompts are made

The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.

Create a capacity planning worksheet for [SERVICE_NAME]. Current peak load and resource usage: [CURRENT_LOAD]. Growth expectations and known events: [GROWTH_EVENTS]. Load test results if any: [LOAD_TEST_RESULTS]. Scaling mechanism: [SCALING_MECHANISM].

Build the worksheet: define the unit of load, measure resource consumption per unit from my data, project load for 3, 6 and 12 months in three scenarios, and compute the resources and cost required with a headroom factor that I choose and justify. Identify the likely first bottleneck and the signal that shows it approaching. Include a load test plan to validate the model: ramp stages, success criteria, and a stop condition to protect shared systems. Present assumptions in a table and highlight the three that drive the answer. Add a review cadence and triggers for an unscheduled review.

Include a small worked example using invented numbers for a service that handles 400 requests per second at 55 percent CPU, showing how to estimate how many instances are needed at double the load with a chosen headroom. Label it clearly as an illustration. Then ask me for the real numbers and list which ones I must measure rather than estimate.

Add a section that turns the projection into a purchasing and scaling decision. For each of the three scenarios show the instance count, the cost per month using placeholders for unit prices, and the date by which the extra capacity must be ordered given the lead time I state. Add a headroom justification table: for each resource (CPU, memory, connections, disk, network) the headroom chosen, the failure it protects against, and the cost of the headroom. Include a section on non-linear effects: queues that grow sharply near saturation, caches that cool after a deploy, and batch jobs that overlap with peak. Finish with a three-line capacity statement suitable for a manager: the current limit, when it is expected to be reached, and the decision needed.

Fill in before use

Optional: type your details and the prompt above updates. Nothing is saved or sent.

Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.

Unit of load: requests per second (rps).
Measured: 400 rps at 55% CPU on 8 instances, so one instance handles about 90 rps at 80% CPU.
Projection (scenario medium): 3 months 520 rps, 6 months 640, 12 months 900.
Instances needed with 30% headroom: 900 / (90 x 0.7) = 14.3, so 15.
Assumptions: linear scaling, no new heavy features.
First bottleneck: the database connection pool at about 700 rps.
Load test: ramp 100 rps every 5 minutes, stop at p95 latency 800 ms.

• capacity