Capacity Planning Skill

Capacity Planning Skill

Produce a complete capacity planning document for a service. Capacity planning is not about predicting the future exactly — it is about understanding current headroom, modelling growth, and ensuring the team takes infrastructure action before a constraint becomes an incident.

A good capacity plan answers: what is running out first, how long before it runs out, what does it cost to fix it, and who decides when to act.

Required Inputs

Ask for these if not already provided:

• Service name and description — what the service does and who depends on it
• Current traffic and usage metrics — requests per second (or per day), active users, data volume — whatever units are most natural for this service
• Current resource utilisation — CPU %, memory %, disk usage, connection pool utilisation, DB query throughput
• Growth rate or projections — historical growth rate, or known upcoming events (product launch, sales cycle, seasonal peak)
• Tech stack and infrastructure — cloud provider, compute type (VMs, containers, serverless), database, caching layer, CDN
• Cost constraints — current infrastructure spend, acceptable cost ceiling, or target cost per unit of traffic

Output Format

---

Capacity Plan: [Service Name]

Service: [Name] | Team: [Team name] Author: [Name] | Last updated: [Date] Planning horizon: [12 months — [Month Year] to [Month Year]] Review cadence: [Quarterly]

---

1. Executive Summary

[3–5 sentences covering: current state, the most critical capacity constraint, the timeline before it becomes a risk, the recommended action, and the cost implication. Written for an engineering manager or VP who needs the key facts without reading the full document.]

Critical finding: [e.g. "The database connection pool will reach 90% utilisation within 6 weeks at current growth. Without action, this will cause request queueing and latency spikes under normal traffic."]

Recommended immediate action: [e.g. "Increase connection pool limit and add a read replica within the next 2 weeks."]

Estimated cost impact: [e.g. "Recommended changes add ~$[X]/month to infrastructure spend."]

---

2. Current Baseline

All metrics are 30-day averages unless noted. Date captured: [Date]

Traffic

Metric	Value	Peak (7-day)	Notes
Requests per second (avg)	[X req/s]	[X req/s]	[Peak time / day of week]
Requests per day	[X M/day]	[X M/day]	—
Active users (DAU/MAU)	[X] / [X]	—	—
[Service-specific metric — e.g. jobs processed/hour]	[X]	[X]	—
[Service-specific metric — e.g. GB ingested/day]	[X GB]	[X GB]	—

Compute

Resource	Current utilisation	Instance type	Count	Notes
CPU (avg)	[X%]	[e.g. c5.2xlarge]	[X]	Peak: [X%]
Memory (avg)	[X%]	—	—	Peak: [X%]
Network egress	[X Mbps]	—	—	—
Container / pod count	[X]	[e.g. 2 vCPU / 4 GB]	—	Auto-scaling range: [X–Y]

Database

Resource	Current utilisation	Spec	Notes
CPU	[X%]	[e.g. db.r5.2xlarge]	Peak: [X%]
Memory	[X%]	[X GB RAM]	—
Storage used	[X GB] of [Y GB] ([Z%])	[X GB provisioned]	Growth: [~X GB/month]
IOPS (avg)	[X] of [Y provisioned]	[Y IOPS]	Peak: [X IOPS]
Connection pool	[X] of [Y max] ([Z%])	Max connections: [Y]	[ORM pool size: X]
Query P99 latency	[X ms]	—	[Slowest query: X]
Read/write ratio	[X%] reads / [Y%] writes	—	—

Cache

Resource	Current utilisation	Spec	Notes
Memory used	[X GB] of [Y GB] ([Z%])	[e.g. cache.r6g.large]	Eviction rate: [X%]
Hit rate	[X%]	—	Miss rate: [Y%]
Connections	[X]	Max: [Y]	—

Storage / Object Store

Resource	Current usage	Growth rate	Notes
[S3 / GCS / Blob]	[X GB / TB]	[~X GB/month]	[Lifecycle policies in place? Y/N]
Disk (if applicable)	[X GB] of [Y GB]	[~X GB/month]	[RAID / EBS type]

Cost Baseline

Component	Current monthly cost	% of total
Compute (app servers)	$[X]	[X%]
Database	$[X]	[X%]
Cache	$[X]	[X%]
Storage	$[X]	[X%]
CDN / bandwidth	$[X]	[X%]
Other ([describe])	$[X]	[X%]
Total	$[X]	100%

Unit economics: $[X] per [1,000 requests / 1,000 users / GB processed]

---

3. Growth Projections

Assumptions

Assumption	Value	Source	Confidence
Monthly traffic growth rate	[X%]	[Historical trend / product forecast]	[High / Medium / Low]
Seasonal peak factor	[+X% in [month(s)]]	[Last year's data / expected launch]	[High / Medium]
Upcoming events	[e.g. Marketing campaign — [Month], expected +[X]% traffic spike]	[Marketing plan]	[Medium]
User growth	[X new users/month]	[Sales pipeline / growth model]	[Medium]
Data growth	[X GB/month]	[Current trend]	[High]

Traffic Forecast

Timeframe	Req/s (avg)	Req/s (peak)	DAU	Data volume (cumulative)
Now (baseline)	[X]	[X]	[X]	[X GB/TB]
+3 months	[X]	[X]	[X]	[X GB/TB]
+6 months	[X]	[X]	[X]	[X GB/TB]
+12 months	[X]	[X]	[X]	[X GB/TB]

Growth formula: [Baseline] × (1 + [monthly rate])^[months] + seasonal adjustment

Capacity Headroom Analysis

When does each resource run out at current utilisation and projected growth?

Resource	Current utilisation	Safe ceiling	Headroom remaining	Months to ceiling
App CPU	[X%]	70%	[X%]	[X months]
App memory	[X%]	80%	[X%]	[X months]
DB CPU	[X%]	70%	[X%]	[X months]
DB storage	[X GB] of [Y GB]	80% = [Z GB]	[X GB]	[X months]
DB IOPS	[X] of [Y]	80% = [Z]	[X IOPS]	[X months]
DB connections	[X] of [Y]	80% = [Z]	[X]	[X months]
Cache memory	[X GB] of [Y GB]	75% = [Z GB]	[X GB]	[X months]
Storage (object)	[X TB]	No hard limit — cost trigger	—	[Cost trigger: $X/month]

Red flags (resources hitting ceiling within 3 months):

• [Resource]: [current]% → ceiling in [X weeks] — Action required
• [Resource]: [current]% → ceiling in [X weeks] — Action required

---

4. Resource Requirements

Compute Requirements

Timeframe	Required instances	Recommended instance type	Auto-scaling range	Notes
Now	[X]	[type]	[min: X, max: Y]	Current configuration
+3 months	[X]	[type]	[min: X, max: Y]	[Any instance type change needed?]
+6 months	[X]	[type or upgrade]	[min: X, max: Y]	[Consider [larger type / horizontal scale]]
+12 months	[X]	[type or upgrade]	[min: X, max: Y]	[State of horizontal vs vertical decision]

Memory headroom target: Maintain ≥30% available memory at average load; ≥20% at peak. CPU headroom target: Maintain ≥30% available CPU at average load; ≥15% at peak.

Database Requirements

Timeframe	Instance type	Storage	IOPS	Read replica	Notes
Now	[type]	[X GB]	[X]	[Y/N]	Current
+3 months	[type]	[X GB]	[X]	[Y/N]	[Upgrade storage / IOPS]
+6 months	[type or upgrade]	[X GB]	[X]	Yes	[Read replica recommended by this point]
+12 months	[type]	[X GB]	[X]	[X replicas]	[Consider sharding / partitioning at this scale]

Storage growth management:

• Current growth: [~X GB/month]
• Storage auto-scaling: [Enabled / Not enabled — enable by [date]]
• Archiving policy: [Records older than X months moved to [cold storage / archive tier]]

Cache Requirements

Timeframe	Node type	Nodes	Memory	Notes
Now	[type]	[X]	[X GB]	Current
+6 months	[type]	[X]	[X GB]	[Scale out or upgrade]
+12 months	[type]	[X]	[X GB]	[Cluster mode if >Y GB required]

---

5. Scaling Strategy

Compute — Horizontal Scaling

Decision: [Horizontal / Vertical / Both]

[State the scaling strategy and the reasoning. E.g. "The application is stateless and CPU-bound; horizontal scaling is preferred. Vertical scaling is a short-term fallback only."]

Auto-scaling configuration:

Scale-out trigger:  CPU > [X%] for [Y minutes] OR memory > [X%] for [Y minutes]
Scale-in trigger:   CPU < [X%] for [Y minutes] AND memory < [X%] for [Y minutes]
Min instances:      [X] (ensures HA across [X] AZs)
Max instances:      [Y] (cost ceiling)
Cooldown period:    [X seconds]
Warmup time:        [X seconds] (time for new instance to be healthy)

Limits of horizontal scaling:

• [e.g. Database connection pool is the current bottleneck — adding more app instances without increasing DB connections will not help]
• [e.g. Session affinity required for WebSocket connections — limits pure stateless scaling]

Database — Read Scaling

Strategy: [Read replica / Connection pooling via PgBouncer / Query caching / None needed yet]

When to add a read replica:

• DB CPU sustained >60% for >30 minutes, OR
• Read query P95 latency >50ms, OR
• Connection pool utilisation >70%

Connection pooling:

• Pooler: [PgBouncer / RDS Proxy / application-level / not configured]
• Pool size: [X connections per app instance × Y instances = Z total]
• Max DB connections: [configured to Z + 20% headroom]

Caching Strategy

Cache policy: [Cache-aside / Write-through / Write-behind] TTL strategy:

Data type	TTL	Invalidation method
[e.g. User profile]	[5 minutes]	[Explicit invalidation on update]
[e.g. Product catalog]	[1 hour]	[TTL expiry — eventual consistency acceptable]
[e.g. Session data]	[24 hours]	[Explicit invalidation on logout]

Cache miss handling: [Describe what happens on a cache miss — does it fall through gracefully or cause a thundering herd risk?]

---

6. Cost Projections

Infrastructure Cost Forecast

Component	Now (monthly)	+3 months	+6 months	+12 months
Compute	$[X]	$[X]	$[X]	$[X]
Database	$[X]	$[X]	$[X]	$[X]
Cache	$[X]	$[X]	$[X]	$[X]
Storage	$[X]	$[X]	$[X]	$[X]
CDN / bandwidth	$[X]	$[X]	$[X]	$[X]
Total	$[X]	$[X]	$[X]	$[X]
MoM growth %	—	[X%]	[X%]	[X%]

Unit economics trend:

Timeframe	Cost per 1k requests	Cost per user/month	Notes
Now	$[X]	$[X]	Baseline
+6 months	$[X]	$[X]	[Improving / worsening — why]
+12 months	$[X]	$[X]	[Target: $X per 1k requests]

Cost optimisation opportunities:

Opportunity	Estimated saving	Effort	Timeline
[e.g. Reserved instances for baseline compute]	$[X/month]	Low	Immediate
[e.g. S3 lifecycle policy — move objects >90 days to Glacier]	$[X/month]	Low	This sprint
[e.g. Right-size [instance] — current is overprovisioned]	$[X/month]	Low	This sprint
[e.g. Optimise top-5 slow queries — reduce DB compute need]	$[X/month]	Medium	Next quarter

---

7. Capacity Triggers and Actions

Define the thresholds that require explicit action — not retrospective fixes after an incident.

Resource	Watch (amber)	Act (red — schedule work)	Emergency (incident risk)
App CPU (sustained avg)	>60%	>70%	>85%
App memory	>70%	>80%	>90%
DB CPU	>55%	>65%	>80%
DB storage	>65%	>75%	>85%
DB connections	>60%	>70%	>85%
Cache memory / eviction	Hit rate <90%	Hit rate <85%	Hit rate <75%
Error rate	>0.5%	>1%	>2%
P99 latency	>2× baseline	>3× baseline	>5× baseline

When a Watch threshold is crossed:

• Engineer who observes it creates a ticket with capacity label
• Ticket reviewed in next sprint planning

When an Act threshold is crossed:

• On-call engineer creates a ticket marked P2
• Tech lead reviews within 24 hours
• Action plan documented and scheduled within 1 sprint

When an Emergency threshold is crossed:

• Treat as a potential incident — page on-call
• Emergency scaling actions taken immediately (see runbook)
• Root cause investigation starts within 2 hours

Emergency scaling runbook: [Link to oncall-runbook for capacity incidents]

---

8. Infrastructure Action Roadmap

Immediate Actions (next 2 weeks)

Action	Owner	Effort	Justification
[e.g. Increase DB connection pool limit to X]	[Name]	[2 hours]	[DB connections at X% — hitting ceiling in X weeks]
[e.g. Enable storage auto-scaling on RDS]	[Name]	[30 min]	[Storage at X% — prevents emergency at X months]
[e.g. Add S3 lifecycle policy for [bucket]]	[Name]	[1 hour]	[Storage growing at $X/month unnecessarily]

This Quarter (within 3 months)

Action	Owner	Effort	Justification
[e.g. Add read replica to production DB]	[Name]	[1 day]	[DB CPU projected to hit 65% in 2 months]
[e.g. Increase max auto-scaling limit from X to Y]	[Name]	[2 hours]	[Current max is too close to expected peak]
[e.g. Configure PgBouncer for connection pooling]	[Name]	[3 days]	[Reduce per-connection overhead; headroom for growth]

Next Quarter (3–6 months)

Action	Owner	Effort	Justification
[e.g. Upgrade DB instance class — [current] → [next]]	[Name]	[2 hours — blue/green]	[DB CPU projected to hit 70% by Q[X]]
[e.g. Implement caching for [high-read endpoint]]	[Name]	[1 week]	[Reduce DB read load by estimated [X%]]
[e.g. Evaluate horizontal DB sharding]	[Name]	[2 weeks (spike)]	[At 12-month projections, single DB hits limits]

Horizon (6–12 months)

Action	Description	Trigger condition
[e.g. Multi-region deployment]	[Active-passive setup in eu-west-2]	[DAU exceeds X or SLA requires 99.99%]
[e.g. Database sharding or migration to distributed DB]	[Evaluate CockroachDB / Vitess]	[Single-node DB projected to hit ceiling]
[e.g. CDN expansion]	[Add PoPs in [region]]	[Latency SLO breached for [geography]]

---

Anti-Patterns

• Do not set capacity trigger thresholds without knowing the baseline — a "CPU > 70%" alert is meaningless if you don't know what normal looks like
• Do not plan only for average traffic — capacity plans that don't model peak load will result in incidents during the events that matter most
• Do not conflate vertical and horizontal scaling — adding more app servers without addressing database connection limits will not resolve the constraint
• Do not present growth projections as certainties — all forecasts have uncertainty; state the confidence level and provide a conservative and optimistic scenario
• Do not defer action items without a named owner and a specific date — a roadmap with no owners is a wish list

Quality Checks

• Every resource has a quantified current utilisation and a projected months-to-ceiling — no hand-waving
• The most critical constraint is called out in the executive summary with a specific timeline
• Growth projections state their assumptions and confidence level — not presented as certainties
• Capacity triggers define amber/red thresholds and name who acts at each level
• Cost projections include unit economics, not just absolute totals
• The infrastructure roadmap has named owners and effort estimates — not just a wish list
• Auto-scaling configuration includes both scale-out AND scale-in triggers, and a min/max range
• Actions are ordered by urgency — immediate items are genuinely immediate, not backlog filler