Collection Runner

The Collection Runner sends a collection’s requests one after another and reports what their tests found. Use it to check a whole API, replay a flow such as sign in, create, read and delete, or run the same requests for every row of a data file.

Start a run

Right-click a collection or folder in the sidebar and choose Run collection or Run folder, or click Run on a collection’s page. The runner opens in a tab with:

• Run sequence: the HTTP requests in the collection’s order, with how many of them are selected. Clear a request’s checkbox to leave it out, or drag a request to send it at another point. Select All (or Deselect All once every request is selected) and Reset change every request at once; Reset also restores the collection’s order. gRPC and WebSocket requests are left out.
• Iterations: how many times the sequence runs.
• Delay: how long to wait between requests, in milliseconds.
• Test data file: a CSV or JSON file whose rows give each iteration its values.
• Advanced settings, described below.

Click Run with the collection’s name to start. Each run sends the requests as they are saved, so save a request’s changes before running it. Variables come from the collection and the environment selected in the tab bar. Values a script sets with pm.variables last for the whole run.

Read the results

The results show how many tests passed and failed, how long the run took and the average response time. Requests are listed by iteration with their status and tests. Passed, Failed, Skipped and Errors narrow the list to the requests with a passing test, a failing test, a skip or an error, and each counts the requests it lists. Console log shows what scripts logged.

Click a request to see its response beside the list: its body, cookies and headers, the request as it was sent, its test results and its console. Use the arrows above the response to see the same request in other iterations. Without a kept response, its test results and console are still there.

While a run is going, a bar shows its progress and the results name the request being sent. Pause waits after the current request and Stop ends the run. Run Again repeats the run; New Run returns to the configuration. Export Results saves the results as JSON; click the saved file’s name to show it in its folder.

Use a data file

A CSV file’s first row names its columns. A JSON file is an array of objects. Each row is one iteration, so choosing a file sets Iterations to its row count; extra iterations reuse the last row.

username,item
ada,book
grace,laptop

Requests use a row’s values as {{username}}, ahead of the environment’s, the collection’s and global values, and scripts read them with pm.iterationData:

pm.test("Signed in as the row's user", () => {
    pm.expect(pm.response.json().user).to.eql(pm.iterationData.get("username"));
});

Choose the next request

A script can change where the run goes next with pm.execution.setNextRequest, as in Postman. Give it a request’s name or ID; the run continues from there after the current request. null ends the iteration.

// Fetch every page before going on.
const body = pm.response.json();
if (body.next) {
    pm.collectionVariables.set("page", body.next);
    pm.execution.setNextRequest("List users");
}

The last choice of a request’s scripts wins. A folder’s run can only go to requests in that folder. When the chosen request is not in the run, the iteration ends and the results say why. Sending one request on its own ignores the choice.

Advanced settings