Azure App Service and Azure Functions can be protected by user authentication with Azure Active Directory due to its authentication/authorization feature (Easy Auth).
However, the authentication methods available in the Apache JMeter™ are limited to Basic, Digest, Kerberos and LDAP authentication.

This tutorial explains how to load testing to Azure App Service with authentication and authorization by Azure AD from Apache JMeter™.

Since this content is authorized by access_token, It's not consider testing for login flow.
Think of it as a test with a login in place.

Advance preparation

There are certain settings that are required for Azure Active Directory to allow access using access_token.

It does not support access by users that require multi-factor authentication.
Turn off multifactor authentication for your test users, or create users that does not require multifactor authentication for testing.

  • Obtaining a client ID of an Azure AD application
    Obtain the client ID of the Azure AD application that the App Service is using for authentication.
    Access the Azure AD Portal( and select the application that is responsible for the App Service authentication in "App Registrations" and note the "Application (client) ID" value in its "Overview".
    [![Get client ID](/conteGet client IDjp/content/images/2020/08/aadauth-clientid_en.png)

  • Obtaining a client secret of an Azure AD application
    Obtain the client secret of the Azure AD application that the App Service is using for authentication.
    If the client secret is not available for this JMeter test, create a new one.

    Note that you will not be able to see this value later.

    In the application "Certificates & secrets", add it from "+ New client secret" and note its value.
    New client secret

  • API permissions
    In the "API permissions" of the application, make sure that the administrator's consent has been granted for "Microsoft Graph > User.Read".
    If not, grant the administrator's consent through "Grant admin content for {ORGANIZATION}".
    API permissions

  • Create a list of users and passwords to login
    Create a list of users and passwords for test login from JMeter as a CSV file.


    Of course, you can also use custom domain account.

  • Install plugin for Microsoft Azure
    Introduce a function plugin to create an Authorization header to make a request to Cosmos DB.

    1. Get the plugin
      Get the latest plugins (jmeter-plugins-functions-azure-{version}.jar) from the following page using a web browser.
    2. Copy the downloaded jar file to $JMETER_HOME/lib/ext
    3. Start or restart JMeter

Create the test plan

  1. Importing a user list from a CSV file
    Add "CSV Data Set Config" to the Test Plan, and read the users and passwords from the CSV file you just created.
    Add > Config Element > CSV Data Set Config
    CSV Data Set Config

    • Filename: CSV file name
    • Variable Names (comma-delimited): username,password
  2. Define Azure AD application information as variables
    Add "User Defined Variables" to the Test Plan and register the information of the Azure AD application that the App Service uses as authentication.
    Add > Config Element > User Defined Variables

    "User Defined Variables" shows two of the same notation, but probably the upper one.
    If you add it, you will see three columns of "Name, Value, Description" in the header of that table.

    User Defined Variables

    • tenent_id: Domain name of Azure AD (ex.,
    • client_id: Client ID of Azure AD application confirmed in advance preparation
    • client_secret: Client secret of Azure AD application confirmed in advance preparation
  3. Add a Thread Group
    Add a "Thread Group" to the Test Plan.
    Add > Threads (Users) > Thread Group
    Enter the appropriate values for the "Number of Threads (users)" and "Loop Count" to access by different users.
    Thread Group

  4. Add a HTTP request sampler to access the page after login
    Add a "HTTP Request" to the Thread Group.
    Add > Sampler > HTTP Request
    HTTP Request

    • Web Server
      • Protocol [http]: https
      • Server Name or IP: Web Apps addresses for load testing with Azure AD authentication (ex.
    • HTTP Request
      • Method: Appropriate value for the request (ex. GET, POST, ...)
      • Path: Path to authenticated page for target of load testing (A page with login user names and other information will output the login user per access)
  5. Add a request header
    Add an "HTTP Header Manager" to the added HTTP Request.
    Add > Config Element > HTTP Header Manager
    HTTP request header

    • Headers Stored in the Header Manager
      • Authorization: Bearer ${__AzAdAccessToken(${tenant_id},password,${client_id},${client_secret},${username},${password})}
  6. Add listeners to see the results of the execution
    Add a "View Results Tree" listener to the Test Plan or thread group.
    Add > Listener > View Results Tree

Execute it and check the results

Run > Start to run the test and select the resulting "HTTP Request" and then select the "Response headers" tab on the "Response data" tab.
Make sure you get "200 OK" here.
Response headers
If you're on a page with a username and other information, you may also want to look at the "Response Body" tab.
Response body
* The Response Body in this image is an example of accessing the following php page.

echo "Hello " . $_SERVER['LOGON_USER'];

By accessing "https://.../.auth/me", you can generally get the details of the logged-in user, but in this procedure it seems that it may return an empty value.

Clean up after the load test

Undo any configuration changes you made for testing, such as removing the client secret if you created it in advance, revoking administrative consent if you gave it to the application's APIs, or removing it if you created a test user.

In this article, I explained how to load testing to Azure App Service with authentication and authorization by Azure AD from Apache JMeter™.

You should be able to load test to Azure Functions with the same settings, so if you're interested, try it out.

Easily build a JMeter multiple remote servers environment in Azure by Load Tester Powered by Apache JMeter™