How to Run a New Load Test

Run a new load test with LoadFocus

It's quite easy to create a new load test for a website or API endpoint using the UI from the New Load Test page.

The page shows the test name at the top, then three numbered steps: What do you want to test?, How much load? and Where from?. Below them, More options holds the optional settings, and the bar at the bottom has the Run test button.

Steps to configure a new load test

1. Choose a Test Name

We automatically generate a test name for your test using the current date and time. It is shown at the top of the page, after New test:. To change it, click the name (or the pencil next to it) and type your own. As the note next to it says, you can rename it any time. Once you type a URL in step 2 and the name is still the automatic one, the website's host name is added in front of it. Letters, digits and dots are kept; any other character in the name becomes an underscore.

2. What do you want to test?

The first numbered step, What do you want to test?, is the place to set the URL endpoint (website, API URI, mobile app URL) under test.

By default, we set the request method to GET (this is the request done when you load a website in a browser). Pick another one from the Method list as your test scenario needs: GET, POST, PUT, PATCH, DELETE, OPTIONS or HEAD.

You can add one or more requests to your load tests with the Add another request button. All Virtual Users will hit the endpoints in sequence, so if you want to reorder the requests, you can easily drag and drop them using the right-hand side arrows button.

3. Headers, body and cookies

Under each URL, click Headers, body, cookies… to show Query Params, Headers, Browser Compatible Headers, Cookies, POST Body, Basic Auth and Extract Variable. While the options are closed, the button shows how many are set, for example "Headers, body, cookies… (2)". See Headers and presets and Using cookies for the second and fourth.

1. Query (URL) parameters

You can add multiple query parameters to your URL by adding key/value pairs with Query Params. The URL box and the list stay in sync, so editing either one updates the other. More in URL query parameters.

This will translate in having a URL: https://www.example.com?key1=value1&key2=value2

4. How much load?

The second step, How much load?, starts with four presets: Quick check (a few users, to check it works), Typical (your everyday traffic), Peak (your busiest moments, to find the limit) and Custom (set your own numbers). Each preset shows the users and total time it sets, scaled to your plan's limits. Click one to fill in the sliders below it; when you change a slider yourself, the matching preset or Custom is selected.

The sliders set the number of Virtual Users, the Duration and the Ramp Up Time that takes the test to the desired load. Open Fine-tune for the Ramp Up Steps and for Run by iterations instead of time, which turns the Duration (s) slider into an Iterations slider: the test then runs a set number of iterations instead of a set time. If a slider is at your plan's limit, a "Need more?" link opens the pricing page. Inspect visually with the chart how your load test will run by viewing the Ramp Up Time and Steps and the Duration of the Load Test. The sentence under the chart sums it up, for example how many users it ramps up to and how long the whole run takes.

Limits during the 7-day trial

A paid plan starts with a 7-day free trial, and during the trial every load test runs at reduced limits regardless of the plan you chose: 50 virtual users, 5-minute tests and 3 test runs per day. That is why the Virtual Users slider stops at 50 on a trial account. A Starter trial is also held at 50 virtual users, but keeps Starter's 20-minute tests. Your plan's full limits (for example 1,500 virtual users on Basic) apply automatically when the trial ends. To unlock them earlier, use Skip the trial and unlock them now in the notice under the slider, or Skip Trial & Pay Now in the limits dialog; billing starts immediately.

The 3 runs a day count across cloud, JMeter and k6 tests. On a trial without a card, runs you made on the Free plan in the last 24 hours count too. Add a card and you get 3 more runs straight away: runs made before you added it no longer count. This works once per trial, and your plan then starts when the trial ends unless you cancel before then. When you reach the daily limit, the limits dialog offers Add a card for 3 more runs today.

5. Where from?

In the third step, Where from?, choose the geographical location from where the virtual clients/users will make requests to your Website/API. On paid plans the list is grouped by region: Americas, Europe, Asia Pacific, and Middle East & Africa. On the Free plan only the first location can be selected; the others are listed under "On paid plans". More details about the locations see here, also whitelisting IPs details can be found here.

6. More options

Under More options you can open these optional rows. Each row shows a short status next to its name, such as On or Off:

• Authorization (applies to every request): set Authorization Type to OAuth 2.0 when the API needs a token (see OAuth 2.0 authorization). The default is No Auth.
• Pass/fail checks: limits each run is graded against (see Pass/Fail Thresholds).
• Run on a schedule: a recurring schedule (see Scheduled load tests).
• Page speed under load: measures Core Web Vitals in a cloud browser while the test runs; it needs a paid Page Speed Monitoring plan.
• Send results to Datadog, New Relic…: sends the results to the integrations you set up.

All of them are optional.

7. Run the test

Click Run test to start the run, or Save draft to keep the configuration without running it. Before the run starts, you confirm it in the Run this test? dialog. When the run finishes, the results page opens with the capacity panel above the charts if the test ramped up.