Skip to content
Documentation

Best Practices

Guides for authoring, running and analysing tests across web automation, API and performance — plus setup, collaboration and release notes.

Suite scalability

As suites grow, lean on reuse, consistent organization, and the cloud to keep them fast and manageable.

Key takeaway: maximize reuse with Step Groups, Elements, and Data Sets, and lean on the cloud's persisted history and concurrency as the suite scales.

Reuse and organize

  • Maximize reuse through Step Groups and Elements to keep large suites DRY and cheap to maintain.
  • Standardize naming and project organization so the suite stays navigable as it grows.
  • Parameterize broadly with Data Sets so a single testcase covers many data variations instead of duplicating testcases.
  • Group related testcases consistently so large suites stay easy to locate, review, and run as a set.

Configure and run at scale

  • Use Test Run Configs per environment and automation type — configs are scoped by automation type and project, so create them under the type you'll run.
  • Prefer Cloud for scale — it persists history and supports concurrency, whereas the Local Agent runs one test at a time and saves nothing.
  • Track runs with releases and tags so you can compare outcomes over time.
  • Keep per-environment configs separate (e.g. staging vs production) rather than editing one config back and forth, so runs stay reproducible.

Respect plan limits

When scaling, stay within your plan's limits (Virtual Users, max execution time, licenses/capacity). Runs queue when capacity is busy or limits are exceeded.

Performance at scale

  • Match the load source to your need — reuse an API/Web testcase for form-driven load, or upload a JMX when you already have a JMeter plan.
  • Choose the run type deliberately — fixed iterations vs time-based; each ignores the other's value.
  • Respect Virtual User and max-execution-time limits — runs queue when capacity is busy and abort when they exceed your plan's limits.
  • Set baselines only from COMPLETED + PASSED runs so comparisons are based on valid results.
  • Scale virtual users gradually — increase load in steps and watch periodic metrics, rather than jumping straight to peak numbers that may queue or hit plan limits.
  • Keep performance configs distinct from Web/API configs — performance configs carry a perf type and are listed only under Performance, so they stay easy to find as the suite grows.

Related guides