Oro Commerce k6 Performance Testing

Oro ships a stock k6 harness (oro/performance-tests) exercising the storefront from outside: browsing, login, listing/search/detail, shopping list, checkout. Plain JavaScript, k6 as a single Go binary or the grafana/k6 image, with named custom metrics you can gate individually.

The Three-Script Pipeline

performance/scripts/, in this fixed order against a cleared cache — step 1 primes what steps 2 and 3 measure, so out of order inflates the numbers.

1. warmingUpTheApp.js — 1 VU / 60s / THRESHOLD_95=3000. Primes OPcache, Doctrine metadata, layout cache, search index.
2. storefrontTests.js — browse flow (home, listing, search, detail, login). Typical 1 VU / 600s / THRESHOLD_95=850.
3. checkoutTest.js — authenticated checkout from a shopping list. Needs SLID, SHIPPINGMETHOD, PAYMENT_METHOD.

Do not reuse the load-run threshold (850 ms) for the warm-up — first-touch is always slower and the run will fail on cold caches.

Environment Variables

All via -e KEY=VALUE. Every script takes BASEURL, USERNAME, PASSWORD, VU, DURATION and THRESHOLD95 (3000 warm-up, 850 load); checkoutTest.js adds SLID, SHIPPINGMETHOD (fixedproduct5) and PAYMENTMETHOD (paymentterm_1). Values and defaults: references/v6.1.md.

Hero Run Command (Docker)

docker run --rm --network host -u "$(id -u):$(id -g)" \
  -v "${PWD}/performance:/home/k6/performance" -w /home/k6/performance \
  grafana/k6:latest run \
  -e BASE_URL="https://oro.docker.local" \
  -e USERNAME="[email protected]" \
  -e PASSWORD="[email protected]" \
  -e VU=1 -e DURATION=60s -e THRESHOLD_95=3000 \
  scripts/warmingUpTheApp.js

--network host lets k6 reach localhost and compose services; -u $(id -u):$(id -g) keeps summary.html from landing root-owned. Every flag: references/docker-invocation.md.

Custom Oro Metrics

The scripts emit 14 named metrics beyond the k6 defaults, so options.thresholds can gate each one rather than a single global p95. List and examples: references/custom-metrics.md.

Reports

handleSummary() writes summary.html via k6-reporter — per-group timings, VU curves, threshold badges — alongside the stdout summary.

Key Pitfalls

1. Load testing without warming up — cold caches inflate mean and p95 for everything. warmingUpTheApp.js at THRESHOLD_95=3000 first, always.
2. Docker without -u $(id -u):$(id -g) — summary.html lands root-owned in the mounted volume; later runs cannot overwrite it and you cannot delete it without sudo.
3. A single global THRESHOLD_95 — a slow listing hides behind a fast login in the aggregate. Threshold the named metrics.
4. Default --network bridge against localhost — the container cannot reach host compose services. Use --network host, or attach it to the compose network.

See Also

In references/: custom-metrics.md (the 14 named metrics) · docker-invocation.md (flag by flag, network modes, ownership) · default-metrics.md (the 9 built-ins and derived KPIs) · install.md · v6.1.md · v7.0.md

• Oro k6 docs
• k6 documentation