Enterprise Cloud Performance Testing Platform

All three run protocol-level load, which measures what your services return under concurrent request volume. For what a browser renders, use Selenium or Playwright on the same platform. Capacity figures are guidance rather than a guarantee, since what each virtual user does decides the real number. Moving an existing managed setup across is a plan import, and the pricing page lists the plan ceilings and metering that decide which tier you land on.

version: 0.2
runson: linux

vars:
  INJECT_TYPE: rampUsers
  USERS: 2000
  RAMP_DURATION: 360
  DURATION: 900

pre:
  - mvn -B clean compile

testSuites:
  - mvn gatling:test -Dgatling.simulationClass=CheckoutSimulation
      -Dgatling.jvmArgs="-DinjectType=${INJECT_TYPE}
      -Dusers=${USERS} -DrampDuration=${RAMP_DURATION}
      -Dduration=${DURATION}"

uploadArtefacts:
  - name: gatling-report
    path:
      - target/gatling/**

jobLabel: [Performance, Gatling]

The sample toolchain is Gatling 3.9.5 with gatling-maven-plugin 4.4.0 on Java 17 and Maven 3.6+. The simulation reads its load shape from system properties, so the same file runs at ten intensities by changing variables. Gatling load testing covers the full configuration, and the HyperExecute performance setup guide has the reference for every field.