Skip to main content

Creating Performance Tests in the UI

Performance tests in Proofarc are powered by Artillery. You can create them from scratch with the Artillery Wizard, or convert an existing API scenario with one click.

Convert from API Scenario (Fastest)​

If you already have a working API scenario, this is the easiest path:

  1. Open any API scenario in the editor
  2. Click the performance icon (or use the menu → Convert to Performance Test)
  3. In the conversion dialog:
    • The scenario summary shows the source endpoints, methods, and base URL
    • Enter a performance test name
    • Choose load configuration:
      • Quick mode — set virtual users (1–100) and duration (10–300s) with sliders
      • Phased mode — select a pattern: Ramp Up, Spike, or Soak
    • Select an SLO preset:
      • Strict — P95 < 200ms, P99 < 500ms, errors < 0.1%
      • Standard — P95 < 500ms, P99 < 1000ms, errors < 1%
      • Relaxed — P95 < 1000ms, P99 < 2000ms, errors < 5%
      • None — no thresholds
    • Optionally check Run immediately
  4. Click Create or Create & Run

Everything carries over: endpoints, request bodies, headers, auth, variable extraction.

Artillery Wizard (From Scratch)​

Create a performance test without an existing API scenario:

Step 1: Target​

  1. In the sidebar, click Performance to expand the section, then click Performance underneath it (the Performance Scenarios page)
  2. Click Create Scenario
  3. Select Project and Environment
  4. Choose a target service (REST_SERVICE targets from the environment)

Step 2: Endpoints​

  1. Select the endpoints to include in the load test
  2. For each endpoint, configure:
    • HTTP method and path
    • Request body (for POST/PUT/PATCH)
    • Headers
  3. Set the test type: Load Test, Stress Test, Endurance Test, or API Test

Step 3: Load Configuration​

Choose a load mode:

Quick mode:

  • Virtual Users slider (1–100)
  • Duration slider (10–300 seconds)

Phased mode — select a pre-built pattern:

  • Ramp Up — gradually increase load
  • Spike — sudden burst of traffic
  • Soak — sustained load over time

Custom mode:

  • Add individual phases manually
  • Per phase: duration, arrival rate, ramp-to target
  • Reorder phases with up/down arrows

Step 4: SLO Thresholds​

Set performance pass/fail criteria:

PresetP95P99Error Rate
Strict< 200ms< 500ms< 0.1%
Standard< 500ms< 1000ms< 1%
Relaxed< 1000ms< 2000ms< 5%

Or set custom thresholds for P95, P99, and error rate.

Step 5: Review & Execute​

Review the configuration summary and click Execute.

Canary Tests (Quick Health Check)​

Run a lightweight performance check directly from any API scenario — no conversion needed:

  1. Open an API scenario
  2. Click the Canary button (speed icon)
  3. In the dialog:
    • Environment auto-selects based on the scenario's target
    • Target service is auto-matched by app tag
    • Credentials — select a credential tag or use the default
    • Set concurrent users (1–50) and requests per user (1–10)
  4. Click Run Canary
  5. Results appear inline:
    • Summary cards: success rate, throughput, P95 latency, failures
    • Pass/Fail verdict based on response times and error rates
    • Per-step breakdown: method, endpoint, latency, success ratio
  6. Click Run Again to re-execute with the same settings

Running a Performance Test​

From the Performance list page:

  1. Click Run on any scenario
  2. In the execution dialog:
    • Review the scenario summary (name, test type, phase count)
    • Select or confirm the target
    • Configure authentication if needed (Bearer, API Key, Basic)
  3. Click Run Test
  4. The test appears in the list with real-time status updates (auto-refreshes every 5 seconds)

Viewing Results​

Click on any completed test to see results:

Summary cards:

  • Total requests with success rate
  • Throughput (requests/second)
  • P95 response time with health indicator (Excellent/Good/Moderate/Slow)
  • Failed request count with error rate

Response Time Details (expandable):

  • Mean, Min, Max response times
  • P95 and P99 percentiles
  • Response time distribution

Status Code Breakdown:

  • Table of HTTP status codes and their counts (200, 404, 500, etc.)

Performance Findings:

  • Auto-generated findings for threshold violations
  • Severity-rated: CRITICAL, HIGH, MEDIUM

Per-Step Metrics (expandable):

  • Individual latency and success rate per API endpoint

Execution History​

View past runs for any performance scenario:

  1. Click the History icon on any scenario in the list
  2. The dialog shows all past executions with:
    • Status (Completed, Failed, Running, Pending)
    • Start time and duration
    • Environment
    • Finding count
  3. Click View Details on any execution to see full results