Peak Season E-Commerce Load and Stress Testing Brief

Role: Senior Performance Engineering Lead specialising in high-throughput retail e-commerce architectures.

Context

• Brand: {{brand_name}}
• Projected traffic: {{projected_rpm}}
• Payment infrastructure: {{payment_gateways}}
• Core catalog database: {{inventory_database_type}}
• Hosting environment: {{cloud_infrastructure}}
• Latency threshold: {{acceptable_p99_latency}}

Task

Author an engineering load and capacity testing brief for {{brand_name}} that details traffic modeling, step-up stress protocols, saturation thresholds, and blast radius isolation testing across {{cloud_infrastructure}} to guarantee checkout stability during promotional flash sales.

Method

1. Analyze historical user journey analytics to build a weighted traffic model incorporating catalog browsing, search indexing, cart additions, and payment processing.
2. Parameterize load generation profiles simulating concurrent shopping spikes up to {{projected_rpm}} across geographically distributed edge nodes.
3. Define performance profiling parameters for {{inventory_database_type}} to detect query lock escalation and inventory reservation race conditions.
4. Design third-party degradation test vectors evaluating circuit breaker behaviors for {{payment_gateways}} under API rate-limiting or timeout conditions.
5. Construct step-up, soak, and spike workload schedules to evaluate auto-scaling velocity within {{cloud_infrastructure}}.
6. Formulate observability checkpoints tracking CPU, memory saturation, p95/p99 response times against {{acceptable_p99_latency}}, and connection pool depletion.
7. Detail automated rollback triggers and infrastructure recovery steps for scenarios where the error budget is exceeded.

Constraints

• MUST establish explicit quantitative pass/fail thresholds based on {{acceptable_p99_latency}}.
• MUST NOT test third-party live endpoints without specifying sandboxed or virtualized mock service layers.
• Load injection scripts must strictly separate authenticated customer paths from anonymous catalog crawler traffic.
• Test datasets must utilize isolated ephemeral records to prevent pollution of production analytics.

Output format

• Executive Workload Profile (table detailing user journey distribution up to {{projected_rpm}})
• Test Execution Phases (numbered list detailing Warmup, Ramp-up, Peak Soak, and Spike phases)
• Metric Baselines and Saturation Limits (table with Metric, Target Baseline, Critical Limit, Alert Trigger)
• Observability and Circuit-Breaker Validation Protocols (max 250 words)

Self-review

1. Verify that the load model reflects the exact {{projected_rpm}} and {{acceptable_p99_latency}} parameters.
2. Confirm that sandbox requirements for {{payment_gateways}} are explicitly addressed.
3. Ensure all four structured sections in the output format are distinctly represented.