In my scenario I had issue on some prod server where after some time we get OOM (OutOfMemory), server is .net IIS, after few quick manual test we was not able to reproduce issue on any our test env. So then Selenium comes to play.
Step one: download.
Selenium IDE is written on top of Firefox - it is Firefox plugin. Download it from:
You need 2 pieces to play: IDE and server. When you install IDE, you'll be ask to restart browser, then you'd see in the tools menu "Selenium IDE".
Step two: record & replay.
You're ready to play - just create your test case, click on "Record" and open the page that you want to play with.
Finish of that is script containing steps that you was doing on page. Replay them to make sure that everything is working. If you'll have error that server is not ready: go to command line and start selenium server: java -jar selenium-server-standalone-{version}.jar
My advice is to do this step couple of time to get familiar how Selenium is working / how is recording steps-clicks, and how actually replay is going through your website - also it looks cool ;-).
When you finish - save it. You also can export it - in my case I export it to jUnit4 format.
When you finish - save it. You also can export it - in my case I export it to jUnit4 format.
Step three: jUnit.
Run from your development IDE (eclipse/netbeans/intelliJ/..).
Create new project in eclipse, then copy body of generated jUnit from previous step. To run test in eclipse you need libraries, you can download them from same page: http://docs.seleniumhq.org/download/. Just add them to your project dependency... and you ready to rock.
You can also use this example maven project:
<project xmlns="http://maven.apache.
<modelVersion>4.0.0</modelVers
<groupId>com.yarenty.uitest</gr
<artifactId>selenium</artifact
<version>0.0.1-SNAPSHOT</versi
<dependencies>
<dependency>
<groupId>org.seleniumhq.
<artifactId>selenium-firefox-
<version>2.38.0</version>
</dependency>
</dependencies>
</project>
How to make concurrent calls? use Executors from my post: http://yarenty.blogspot.com/2012/01/template-multithreading-solution-using.html
public void start(){
ExecutorService executor = Executors.newFixedThreadPool(CONCURRENT_USERS);
List<Future<Long>> list = new ArrayList<Future<Long>>();
for (int i = 0; i < NUMBER_OF_USER_SESSIONS*CONCURRENT_USERS; i++) {
Callable<Long> worker = new FirefoxWorker(i);
Future<Long> submit = executor.submit(worker);
list.add(submit);
}
// now retrieve the result
for (Future<Long> future : list) {
try {
future.get();
} catch (InterruptedException e) {
e.printStackTrace();
} catch (ExecutionException e) {
e.printStackTrace();
}
}
executor.shutdown();
}
CONCURRENT_USERS - is what is the number of users accessing webpages at single time.
NUMBER_OF_USER_SESSIONS - is how many interactions do you want to have.
Overall number of "visits" is multiplicity of both above.
In my solution I created user_id adding to standard user "i" - next number of execution. You may think about different solution - Hopefully I can present you how to incorporate Disruptor with defined handlers, ... soon ;-)
