All You Need To Know For Selenium Testing On The Cloud

Selenium: A Popular Alternative

Selenium is a very popular test framework using which you can create scalable test cases for automated browser testing or cross browser testing. It is an open-source framework that automates a browser to behave like a user is expected to interact with your website. You can write test scripts in many languages, including but not limited to Java, Python, PHP, Ruby, C#, etc.

A Detailed Study on What is Selenium Grid?

There are innumerable factors that can be attributed to the popularity of the Selenium framework; the major ones are mentioned below:

• Selenium is open-source, and its usage does not involve any overhead costs.
• Popular programming languages like C#, Python, JavaScript, Ruby on Rails, etc. can be used seamlessly with the Selenium framework.
• Selenium has a vast support community to help you out of a sticky situation. This community is growing every single day.
• WebDrivers for popular web browsers like Chrome, Firefox, Microsoft Edge, etc. can be downloaded to automate GUI tests.
• Popular frameworks like Hypothesis, Robot Framework, etc. (for Python) can perform testing using a wide range of input values.

Merits Of Selenium Testing On The Cloud

Automated browser testing on local infrastructure is neither economical nor scalable. Even if you are willing to invest heavily in the overall infrastructure, you will not achieve the same quality of throughput, test performance, parallelism, test coverage, etc. that you are likely to accomplish with cross browser testing in Selenium.

By Sathwik Prabhu

Below are some of the most important benefits of switching to Selenium testing on the cloud:

• Achieve better browser coverage, platform coverage, and test coverage by executing automation tests on cloud-based cross browser testing platforms.
• Scalable and cost-effective as you pay for the resources that you use. It works on a SAAS (Software as a service) model.
• Execute automation tests at a faster pace by leveraging the power of parallel testing.
• Use browser capabilities to test on new as well as outdated/old web browsers, operating systems, and devices.
• Use popular programming languages like C#, JavaScript, Ruby on Rails, Python, etc. along with the Selenium test framework, i.e. remote Selenium WebDriver for scalable development test cases/test suites.
• Keep track of the testing process’s evolution and take advantage of bug-tracking, communication, and project management tools for tracking & timely communication with other team members.
• Leverage the advantage of CI/CD (Continuous Integration/Continuous Delivery) tools like Jenkins, TeamCity, Travis, Bamboo, etc. to achieve automation of the build, package, and testing processes.

Selenium Testing On The Cloud In Action

Now that you are all set with a TestMu AI account, it is time to execute cross browser testing in Selenium. For now, let us look into a Python example that demonstrates Selenium testing on the cloud using remote Selenium Grid on TestMu AI.

Serial Testing

Serial testing is a practice of executing tests serially, i.e. tests are executed one after another. Once the first test is completed, the next test will be executed. The downside of using serial testing is that it increases the test execution time since the Selenium testing framework and cloud infrastructure capabilities are not used to the fullest.

Problem Statement: Below is the breakdown of the things that we plan to achieve with this test:

• Open DuckDuckGo in the Firefox browser.
• Locate the Search Box.
• Enter a search query, i.e. TestMu AI.
• Execute the search operation.
• Free up the resources.

Implementation:

# Python example demonstrates Selenium testing on the Cloud using LambdaTest platform
import time
import unittest
import selenium
import urllib3
import time
import warnings
from time import sleep
from selenium import webdriver
 
test_url = "https://www.duckduckgo.com"
 
# Desired browser capabilities generated using LambdaTest Capabilities Generator
# https://www.lambdatest.com/capabilities-generator/
 
desired_capabilities = {
        "build" : "Selenium testing on the Cloud using LambdaTest",
        "name" : "Selenium testing on the Cloud using LambdaTest",
        "platform" : "Windows 10",
        "browserName" : "Chrome",
        "version" : "72.0"
    }
 
class SeachTest(unittest.TestCase):
    def setUp(self):
        print("Start - SetUp")
 
        # Ignore ResourceWarning related warnings
        warnings.filterwarnings(action="ignore", message="unclosed", category=ResourceWarning)
 
        urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
        user_name = "user-name"
        app_key = "app_key"
 
        # Username and Access Token to use the LambdaTest platform
        remote_url = "https://" + user_name + ":" + app_key + "@hub.lambdatest.com/wd/hub"
        self.driver = webdriver.Remote(command_executor=remote_url, desired_capabilities=desired_capabilities)
 
        print("End - SetUp")
 
    def test_selenium_on_cloud(self):
        print("Start - test_selenium_on_cloud")
        self.driver.maximize_window()
        self.driver.get(test_url)
 
        # time.sleep(5)
        sleep(5)
 
        # Locate the element using the Inspect tool of the browser
        elem = self.driver.find_element_by_xpath("//*[@id='search_form_input_homepage']")
        elem.send_keys("Lambdatest")
 
        # Execute the search
        elem.submit()
 
        # Sleep for 10 seconds in order to see the results
        sleep(5)
 
        print("End - test_selenium_on_cloud")
        
    def tearDown(self):
        print("Start - tearDown")
        # Close the browser.
        self.driver.quit()
 
        print("End - tearDown")
 
if __name__ == '__main__':
    unittest.main()

A combination of the user-name and the access-token is used for logging on to the platform.

user_name = "user-name"
app_key = "app_key"
 
# Username and Access Token to use the LambdaTest platform
remote_url = "https://" + user_name + ":" + app_key + "@hub.lambdatest.com/wd/hub"

The browser and platform capabilities are generated using the TestMu AI Capabilities Generator. You should select the programming language used for implementation and then the browser specifications. In our case, it is Python.

desired_capabilities = {
        "build" : "Selenium testing on the Cloud",
        "name" : "Selenium testing on the Cloud using LambdaTest",
        "platform" : "Windows 10",
        "browserName" : "Chrome",
        "version" : "72.0"
}
 
………………………………………………………………………
………………………………………………………………………
self.driver = webdriver.Remote(command_executor=remote_url, desired_capabilities=desired_capabilities)

Using the Inspect Tool of the browser, we find the details of the Web Element. In our case, it is the Search Box on the DuckDuckGo homepage.

We use the element’s property of XPath in Selenium, and the same would be used in the code.

# Locate the element using the Inspect tool of the browser
elem = self.driver.find_element_by_xpath("//*[@id='search_form_input_homepage']")
elem.send_keys("Lambdatest")

For the demonstration of Selenium testing on the cloud, we have made use of the unittest framework (in Python) along with the Selenium test suite. The execution is done by invoking the python command from the terminal.

python <file-name.py>

You can check this automation test’s status on this page by navigating to the corresponding Test ID & Build ID.

Parallel Testing

Parallel testing in Selenium involves running a test automation suite or test cases in parallel. The test cases are not executed in a serialized manner, which leads to a reduction of test execution time. A combination of test suites is executed in parallel on multiple servers, with each server running one test at a time.

For speedy execution on TestMu AI, you can use parallel testing where concurrent sessions are triggered on different servers. As shown in the snapshot below, we can invoke two concurrent sessions. The Pricing Plans have more information about the availability of parallel execution for your test cases.

Problem Statement: Below is the breakdown of the things that we plan to achieve in the two tests:

Both the tests should execute in parallel, i.e. the number of concurrent sessions on the TestMu AI server should be 2/2 in this case.

Implementation:

As the only difference between the two tests is the browser on which the tests are executed, we generate the corresponding browser capabilities for Chrome & IE browser using the TestMu AI Capabilities Generator.

Test 1: Browser Capabilities [Chrome]

desired_capabilities = {
"build" : "Parallel Testing - Selenium testing on the Cloud [Chrome]",
"name" : "Parallel Testing - Selenium testing on the Cloud [Chrome]",
"platform" : "Windows 10",
"browserName" : "Chrome",
"version" : "72.0"
}

Test 2: Browser Capabilities [IE]

desired_capabilities = {
"build" : "Parallel Testing - Selenium testing on the Cloud [IE]",
"name" : "Parallel Testing - Selenium testing on the Cloud [IE]",
"platform" : "Windows 7",
"browserName" : "Internet Explorer",
"version" : "11.0",
"ie.compatibility" : 11001
}

The rest of the implementation remains the same as in the case of serial testing. Shown below is the implementation for the IE browser.

# Python example demonstrates Selenium testing on the Cloud using LambdaTest platform
import time
import unittest
import selenium
import urllib3
import time
import warnings
from time import sleep
from selenium import webdriver
 
test_url = "https://www.duckduckgo.com"
 
# Desired browser capabilities generated using LambdaTest Capabilities Generator
# https://www.lambdatest.com/capabilities-generator/
 
desired_capabilities = {
    "build" : "Parallel Testing - Selenium testing on the Cloud [IE]",
    "name" : "Parallel Testing - Selenium testing on the Cloud [ChromeIE]",
    "platform" : "Windows 7",
    "browserName" : "Internet Explorer",
    "version" : "11.0",
    "ie.compatibility" : 11001
}
 
class SeachTest(unittest.TestCase):
    def setUp(self):
        print("Start - SetUp")
 
        # Ignore ResourceWarning related warnings
        warnings.filterwarnings(action="ignore", message="unclosed", category=ResourceWarning)
 
        urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
        # user_name = "user-name"
        # app_key = "app_key"
        user_name = "[email protected]"
        app_key = "fbI6kxucn5iRzwt5GWYiNvaPb4Olu9R8lwBsXWTSaIOebXn4x9"
 
        # Username and Access Token to use the LambdaTest platform
        remote_url = "https://" + user_name + ":" + app_key + "@hub.lambdatest.com/wd/hub"
        self.driver = webdriver.Remote(command_executor=remote_url, desired_capabilities=desired_capabilities)
 
        print("End - SetUp")
 
    def test_selenium_on_cloud(self):
        print("Start - test_selenium_on_cloud")
        self.driver.maximize_window()
        self.driver.get(test_url)
 
        # time.sleep(5)
        sleep(5)
 
        # Locate the element using the Inspect tool of the browser
        elem = self.driver.find_element_by_xpath("//*[@id='search_form_input_homepage']")
        elem.send_keys("Lambdatest")
 
        # Execute the search
        elem.submit()
 
        # Sleep for 10 seconds in order to see the results
        sleep(5)
 
        print("End - test_selenium_on_cloud")
        
    def tearDown(self):
        print("Start - tearDown")
        # Close the browser.
        self.driver.quit()
 
        print("End - tearDown")
 
if __name__ == '__main__':
    unittest.main()

For Parallel execution, both the files [one for testing on Chrome and other for testing on IE] are invoked from the terminal at the same point.

As seen from the execution snapshot on TestMu AI, tests are executed in parallel due to which the overall execution time is reduced. The more the number of concurrent sessions, the lesser will be the execution time.