Below are possible ways to run Selenium test-cases from JMeter:

• using JUnit Request Sampler;
• using BeanShell Sampler;
• using JSR223 Sampler + Groovy.

JUnit Request Sampler

Running Selenium tests this way maybe useful if you want to re-use already automated (Java) Selenium scenarios instead of re-writing JS-scripts for WebDriver Sampler.

Selenium RC

---

1. Prepare Selenium test project and setup.

   1.1. Download Selenium Java client libraries and put selenium-java-${version}.jar to JMeter classpath, e.g. %JMETER_HOME%/lib/.
   1.2. Selenium server should be up and listening:

   java -jar selenium-server-standalone-${version}.jar
   

   1.3. Export Selenium test-plan as .jar and save it to %JMETER_HOME%/lib/junit/.

   NOTE: Your test class should extend TestCase or SeleneseTestCase to allow JMeter pick up this test plan, test case's name should start with "test").
   NOTE: By default SeleneseTestCase extends JUnit 3.x TestCase, also SeleneseTestCase expects external Selenium server to be running.

2. Configure JUnit Request sampler

   2.1. In JMeter test-plan add JUnit Request sampler.
   Set Class Name according to one from the Selenium test plan.
   Set Test Method to test that is about to run.
   Leave other parameters by default.

   

   JUnit 3.x vs. 4.x
   JUnit Request Sampler can process both JUnit3- and JUnit4-style classes and methods. To set Sampler to search for JUnit 4 tests (@Test annotations) check Search for Junit4 annotations (instead of JUnit 3) checkbox in settings above.
   The following JUnit4 annotations are recognized:

   @Test - used to find test methods and classes. The "expected" and "timeout" attributes are supported.
   @Before - treated the same as setUp() in JUnit3
   @After - treated the same as tearDown() in JUnit3
   @BeforeClass, @AfterClass - treated as test methods so they can be run independently as required

3. You are ready to start your Selenium test with JMeter.

Java code for JUnit Request sampler:

JUnit 3.x

package com.example.tests;

import com.thoughtworks.selenium.*;

public class selenium extends SeleneseTestCase {

    private static Selenium selenium;

    public void setUp() throws Exception {
        selenium = new DefaultSelenium("localhost", 4444, "*firefox", "http://www.google.com/");
        selenium.start();
        selenium.windowMaximize();
    }

    public void testSelenium() throws Exception {
        selenium.open("/");
        selenium.waitForPageToLoad("30000");
        Assert.assertEquals("Google", selenium.getTitle());
    }

    public void tearDown() throws Exception {
        selenium.close();
    }
}

JUnit 4.x

Test script written in JUnit 4 uses JUnit annotations:

package com.example.tests;

import com.thoughtworks.selenium.*;

import org.junit.After;
import org.junit.Assert;
import org.junit.Before;
import org.junit.Test;

public class selenium extends SeleneseTestCase {

    private static Selenium selenium;

    @Before
    public void setUp() throws Exception {
        selenium = new DefaultSelenium("localhost", 4444, "*firefox", "http://www.google.com/");
        selenium.start();
        selenium.windowMaximize();
    }

    @Test
    public void testSelenium() throws Exception {
        selenium.open("/");
        selenium.waitForPageToLoad("30000");
        Assert.assertEquals("Google", selenium.getTitle());
    }

    @After
    public void tearDown() throws Exception {
        selenium.stop();
    }
}

Selenium WebDriver

---

This case is an alternative to WebDriver Sampler mentioned in another answer below.

Prerequisites

The only difference with Selenium RC case is Selenium setup preparation:

1.1. Download and put selenium-server-standalone-${version}.jar to JMeter classpath, e.g. %JMETER_HOME%/lib/.
NOTE: There is no need to start the Selenium server.

All the other steps are the same as in the scenario described above.

package org.openqa.selenium.example;

import junit.framework.TestCase;

import org.junit.Before;
import org.junit.Test;
import org.junit.After;
import org.openqa.selenium.*;
import org.openqa.selenium.WebDriver;
import org.openqa.selenium.WebElement;
import org.openqa.selenium.firefox.FirefoxDriver;
import org.openqa.selenium.firefox.FirefoxProfile;

public class selenium extends TestCase {
    public static WebDriver driver;

    @Before
    public void setUp() {
        FirefoxProfile profile = new FirefoxProfile();
        driver = new FirefoxDriver(profile);
    }

    @Test
    public void testSelenium() throws Exception {
        driver.get("http://www.google.com/");
        Assert.assertEquals("Google", driver.getTitle());
    }

    @After
    public void tearDown() {
        driver.quit();
    }
}

---

Upd.

Another good points and step-by-step guides to use Selenium + JUnit + JMeter bundle:

• Integrating Selenium with JMeter for Load Testing
• Integrating Jmeter with Selenium Code
• Performance testing with Selenium and JMeter
• Running Selenium tests under JMeter
• How to integrate a JUnit4 – Webdriver test into JMeter

---

BeanShell Sampler

In this case selenium test-scenario is executed directly in JMeter's BeanShell Sampler.

1. Selenium setup preparation is completely identical to described above cases: download Selenium libraries, put to JMeter classpath, start Selenium server (in case of Selenium RC).
2. Put your selenium test-scenario into BeanShell Sampler:

Selenium RC

import com.thoughtworks.selenium.*;
import java.util.regex.Pattern;

Boolean result = true;

try {
    selenium = new DefaultSelenium("localhost", 4444, "*iexplore", "http://www.google.com/");
    selenium.start();
    selenium.windowMaximize();

    selenium.open("/");
    selenium.waitForPageToLoad("30000");  

    if (!selenium.isTextPresent("Google")) result = false;
} catch (Exception ex) {
    ex.printStackTrace();
    IsSuccess = false;
    ResponseCode = "500";
    ResponseMessage = ex.getMessage();
} finally {
    selenium.stop();
}

IsSuccess = result;
return result;

Selenium WebDriver

import org.openqa.selenium.By;
import org.openqa.selenium.WebDriver;
import org.openqa.selenium.WebElement;
import org.openqa.selenium.htmlunit.HtmlUnitDriver;

Boolean result = true;

try {
    driver = new HtmlUnitDriver();
    driver.setJavascriptEnabled(true);

    driver.get("http://www.google.com/");

    if (!driver.getTitle().contains("Google")) result = false;
} catch (Exception ex) {
    ex.printStackTrace();
    IsSuccess = false;
    ResponseCode = "500";
    ResponseMessage = ex.getMessage();
} finally {
    driver.quit();
}

IsSuccess = result;
return result;

---

JSR223 Sampler + Groovy

In this case selenium test-scenario is executed via JSR223 Sampler + Groovy.
For performance considerations this approach seems to be more preferable than using BeanShell Sampler described above.

1. Selenium setup preparation is completely identical to described above cases: download Selenium libraries, put to JMeter classpath, start Selenium server (in case of Selenium RC).

2. Add Groovy support for JSR223 Sampler:

   2.1. download latest Groovy binary distribution;
   2.2. copy groovy-all-${VERSION}.jar from “embeddable” folder of distribution and drop it to %JMETER_HOME%/lib/;
   2.3. restart JMeter.

3. Configure JSR233 Sampler:

   3.1. add JSR233 Sampler to Thread Group;
   3.2. set Script Language to groovy in sampler's settings;
   3.3. put your selenium test-scenario into Script section (Java code will be accepted):

Selenium RC

import com.thoughtworks.selenium.*;
import java.util.regex.Pattern;

Boolean result = true;

try {
    selenium = new DefaultSelenium("localhost", 4444, "*iexplore", "http://www.google.com/");
    selenium.start();
    selenium.windowMaximize();

    selenium.open("/");
    selenium.waitForPageToLoad("30000");      

    if (!selenium.isTextPresent("Google")) result = false;
} catch (Exception ex) {
    ex.printStackTrace();
    log.error(ex.getMessage());
     SampleResult.setSuccessful(false);
     SampleResult.setResponseCode("500");
     SampleResult.setResponseMessage(ex.getMessage());
} finally {
    selenium.stop();
}

SampleResult.setSuccessful(result);
return result;

Selenium WebDriver

import org.openqa.selenium.By;
import org.openqa.selenium.WebDriver;
import org.openqa.selenium.WebElement;
import org.openqa.selenium.htmlunit.HtmlUnitDriver;

Boolean result = true;

try {
    driver = new HtmlUnitDriver();
    driver.setJavascriptEnabled(true);

    driver.get("http://www.google.com/");

    if (!driver.getTitle().contains("Google")) result = false;
} catch (Exception ex) {
    ex.printStackTrace();
    log.error(ex.getMessage());
     SampleResult.setSuccessful(false);
     SampleResult.setResponseCode("500");
     SampleResult.setResponseMessage(ex.getMessage());
} finally {
    driver.quit();
}

SampleResult.setSuccessful(result);
return result;

---

Common notes for BeanShell / JSR223 Sampler cases:

• Use external .bsh / .groovy files with test-scenario (Script file field) instead of using Beanshell / Groovy code directly in sampler for intensive testing.
• Since BeanShell / JSR233 Samplers have access to JMeter's variables you can set test (= sampler execution) status directly in test-scenario (via e.g. IsSuccess = STATUS or SampleResult.setSuccessful(STATUS), see code above), without using Response Assertion.

You can also check pageloaded using following code

IWait<IWebDriver> wait = new OpenQA.Selenium.Support.UI.WebDriverWait(driver, TimeSpan.FromSeconds(30.00));

 wait.Until(driver1 => ((IJavaScriptExecutor)driver).ExecuteScript("return document.readyState").Equals("complete"));

The API documentation for findElements() says:

When using xpath be aware that webdriver follows standard conventions: a search prefixed with "//" will search the entire document, not just the children of this current node. Use ".//" to limit your search to the children of this WebElement.

So you should be able to do something like:

WebElement foo = driver.findElement(By.xpath("//*[@title='" + title + "']"));
WebElement bar = foo.findElement(By.xpath(".//img"));

The leading dot in the second xpath makes all the difference, as it will search only for child elements of foo.

A few quick searches led me to this:

http://blogs.steeplesoft.com/posts/2012/grabbing-screenshots-of-failed-selenium-tests.html

Basically, he recommends creating a JUnit4 Rule that wraps the test Statement in a try/catch block in which he calls:

imageFileOutputStream.write(
    ((TakesScreenshot) driver).getScreenshotAs(OutputType.BYTES));

Does that work for your problem?

The Selenium client bindings will try to locate the geckodriver executable from the system PATH. You will need to add the directory containing the executable to the system path.

• On Unix systems you can do the following to append it to your system’s search path, if you’re using a bash-compatible shell:

  export PATH=$PATH:/path/to/geckodriver
  

• On Windows you need to update the Path system variable to add the full directory path to the executable. The principle is the same as on Unix.

All below configuration for launching latest firefox using any programming language binding is applicable for Selenium2 to enable Marionette explicitly. With Selenium 3.0 and later, you shouldn't need to do anything to use Marionette, as it's enabled by default.

To use Marionette in your tests you will need to update your desired capabilities to use it.

Java :

As exception is clearly saying you need to download latest geckodriver.exe from here and set downloaded geckodriver.exe path where it's exists in your computer as system property with with variable webdriver.gecko.driver before initiating marionette driver and launching firefox as below :-

//if you didn't update the Path system variable to add the full directory path to the executable as above mentioned then doing this directly through code
System.setProperty("webdriver.gecko.driver", "path/to/geckodriver.exe");

//Now you can Initialize marionette driver to launch firefox
DesiredCapabilities capabilities = DesiredCapabilities.firefox();
capabilities.setCapability("marionette", true);
WebDriver driver = new MarionetteDriver(capabilities); 

And for Selenium3 use as :-

WebDriver driver = new FirefoxDriver();

If you're still in trouble follow this link as well which would help you to solving your problem

.NET :

var driver = new FirefoxDriver(new FirefoxOptions());

Python :

from selenium import webdriver
from selenium.webdriver.common.desired_capabilities import DesiredCapabilities

caps = DesiredCapabilities.FIREFOX

# Tell the Python bindings to use Marionette.
# This will not be necessary in the future,
# when Selenium will auto-detect what remote end
# it is talking to.
caps["marionette"] = True

# Path to Firefox DevEdition or Nightly.
# Firefox 47 (stable) is currently not supported,
# and may give you a suboptimal experience.
#
# On Mac OS you must point to the binary executable
# inside the application package, such as
# /Applications/FirefoxNightly.app/Contents/MacOS/firefox-bin
caps["binary"] = "/usr/bin/firefox"

driver = webdriver.Firefox(capabilities=caps)

Ruby :

# Selenium 3 uses Marionette by default when firefox is specified
# Set Marionette in Selenium 2 by directly passing marionette: true
# You might need to specify an alternate path for the desired version of Firefox

Selenium::WebDriver::Firefox::Binary.path = "/path/to/firefox"
driver = Selenium::WebDriver.for :firefox, marionette: true

JavaScript (Node.js) :

const webdriver = require('selenium-webdriver');
const Capabilities = require('selenium-webdriver/lib/capabilities').Capabilities;

var capabilities = Capabilities.firefox();

// Tell the Node.js bindings to use Marionette.
// This will not be necessary in the future,
// when Selenium will auto-detect what remote end
// it is talking to.
capabilities.set('marionette', true);

var driver = new webdriver.Builder().withCapabilities(capabilities).build();

Using RemoteWebDriver

If you want to use RemoteWebDriver in any language, this will allow you to use Marionette in Selenium Grid.

Python:

caps = DesiredCapabilities.FIREFOX

# Tell the Python bindings to use Marionette.
# This will not be necessary in the future,
# when Selenium will auto-detect what remote end
# it is talking to.
caps["marionette"] = True

driver = webdriver.Firefox(capabilities=caps)

Ruby :

# Selenium 3 uses Marionette by default when firefox is specified
# Set Marionette in Selenium 2 by using the Capabilities class
# You might need to specify an alternate path for the desired version of Firefox

caps = Selenium::WebDriver::Remote::Capabilities.firefox marionette: true, firefox_binary: "/path/to/firefox"
driver = Selenium::WebDriver.for :remote, desired_capabilities: caps

Java :

DesiredCapabilities capabilities = DesiredCapabilities.firefox();

// Tell the Java bindings to use Marionette.
// This will not be necessary in the future,
// when Selenium will auto-detect what remote end
// it is talking to.
capabilities.setCapability("marionette", true);

WebDriver driver = new RemoteWebDriver(capabilities); 

.NET

DesiredCapabilities capabilities = DesiredCapabilities.Firefox();

// Tell the .NET bindings to use Marionette.
// This will not be necessary in the future,
// when Selenium will auto-detect what remote end
// it is talking to.
capabilities.SetCapability("marionette", true);

var driver = new RemoteWebDriver(capabilities); 

Note : Just like the other drivers available to Selenium from other browser vendors, Mozilla has released now an executable that will run alongside the browser. Follow this for more details.

You can download latest geckodriver executable to support latest firefox from here

Use .getOptions() method and store them in a list .Then find its size.

Select se = new Select(driver.findElement(By.id("select drop down locator")));

List<WebElement> l = se.getOptions();
l.size();

-Ajay

Just for anyone else who's looking for an answer in Ruby, Python, and C# bindings (Selenium 2.33.0).

Note that the actual keys to send depend on your OS. For example, Mac uses CMD + T, instead of Ctrl + T.

Ruby

require 'selenium-webdriver'

driver = Selenium::WebDriver.for :firefox
driver.get('http://stackoverflow.com/')

body = driver.find_element(:tag_name => 'body')
body.send_keys(:control, 't')

driver.quit

Python

from selenium import webdriver
from selenium.webdriver.common.keys import Keys

driver = webdriver.Firefox()
driver.get("http://stackoverflow.com/")

body = driver.find_element_by_tag_name("body")
body.send_keys(Keys.CONTROL + 't')

driver.close()

C#

using OpenQA.Selenium;
using OpenQA.Selenium.Firefox;

namespace StackOverflowTests {

    class OpenNewTab {

        static void Main(string[] args) {

            IWebDriver driver = new FirefoxDriver();
            driver.Navigate().GoToUrl("http://stackoverflow.com/");

            IWebElement body = driver.FindElement(By.TagName("body"));
            body.SendKeys(Keys.Control + 't');

            driver.Quit();
        }
    }
}

Usually Modular Windows are part of the same DOM, unlike javascripts alerts. Only thing that sets them apart from rest of the page is that they are in different frame.

Try to see if this Modular Window lies inside a frame or iframe tag. If any of the parent is frame or iframe then you will have to change the context to that frame before you can performa any action on the Modal Window.

So find the frame do a driver.switchTo().frame() and then perform the action on the element you want to. Once the action is done, which would most probably bring you back to the main page. Use driver.switchTo().defaultContent() to bring focus back to main page.

This SO question will be helpful.

If this does not work it would be helpful to have a look at the page or its HTML.

Here is the working code block to login into your Gmail account through a valid set of credentials-

System.setProperty("webdriver.gecko.driver","C:\\your_directory\\geckodriver.exe");
WebDriver driver = new FirefoxDriver();
driver.manage().window().maximize();
String url = "https://accounts.google.com/signin";
driver.get(url);
driver.manage().timeouts().implicitlyWait(10, TimeUnit.SECONDS); 
WebElement email_phone = driver.findElement(By.xpath("//input[@id='identifierId']"));
email_phone.sendKeys("your_email_phone");
driver.findElement(By.id("identifierNext")).click();
WebElement password = driver.findElement(By.xpath("//input[@name='password']"));
WebDriverWait wait = new WebDriverWait(driver, 20);
wait.until(ExpectedConditions.elementToBeClickable(password));
password.sendKeys("your_password");
driver.findElement(By.id("passwordNext")).click();

---

Update(5-Jan-2020)

Optimizing the above code block and adding a couple of arguments you can use:

public class browserAppDemo 
{
    public static void main(String[] args) throws Exception 
    {
        System.setProperty("webdriver.chrome.driver", "C:\\Utility\\BrowserDrivers\\chromedriver.exe");
        ChromeOptions options = new ChromeOptions();
        options.addArguments("start-maximized");
        options.setExperimentalOption("useAutomationExtension", false);
        options.setExperimentalOption("excludeSwitches", Collections.singletonList("enable-automation"));
        WebDriver driver =  new ChromeDriver(options); 
        driver.get("https://accounts.google.com/signin")
        new WebDriverWait(driver, 10).until(ExpectedConditions.elementToBeClickable(By.xpath("//input[@id='identifierId']"))).sendKeys("emailID");
        driver.findElement(By.id("identifierNext")).click();
        new WebDriverWait(driver, 10).until(ExpectedConditions.elementToBeClickable(By.xpath("//input[@name='password']"))).sendKeys("password");
        driver.findElement(By.id("passwordNext")).click();
    }
}