Spike test

Reliability and performance tests

1 min

Intro

Spike test evaluates system behavior under a sudden, sharp increase in load followed by an immediate decrease. The goal is to measure resilience and recovery after an abrupt surge in traffic.

Easy explanation

Checking how a bridge reacts when an unexpected rush of cars crosses it at once.

Scheme

Spike test scheme

Info

To better understand the terms in testing and tool k6 see vocabulary.

Key characteristics

Sudden sharp load increase: Quickly ramps up users far beyond normal load.
Short duration: Spike lasts only briefly.
Recovery focus: Observes how quickly and gracefully the system recovers after the spike.
Resilience check: Ensures no lingering effects on performance or stability.

Practical examples

How does the system handle an instant increase from 100 to 1000 users?
Does performance degrade temporarily or permanently after the spike?

Users and time

Number of virtual users:

Simulates rapid jump to 2–3x of the load. Example: If load is 500 users — spike test with 1000–1500 users.
Execution time:

Short test — typically 5 to 15 minutes, spike itself lasting 1–2 minutes.

Examples with tool k6

Simple example

script.js

import http from 'k6/http';
import { check, sleep } from 'k6';

export const options = {
  stages: [
    { duration: '30s', target: 0 }, // idle
    { duration: '30s', target: 1000 }, // sudden spike to 1000 users
    { duration: '1m', target: 1000 }, // hold for 1 minute
    { duration: '30s', target: 0 }, // drop to 0 users
  ],
  thresholds: {
    http_req_failed: ['rate<0.01'], // <1% errors
    http_req_duration: ['p(95)<1000'], // 95% requests < 1s
  },
};

export default function () {
  const response = http.get('https://api.example.com/products');

  check(response, {
    'status is 200': (res) => res.status === 200,
  });

  sleep(1);
}

Deep dive handbook