I'm building VS.NEt 2010 Load Test solution.
Everything works really good except one thing. When I record .webtest script - it grabs the site domain name, like so:
http://test1/page1
http://test1/page2
So the test1 is hardcoded in the script.
What I would like to do is to run same load test again different test environment - the goal is to compare two environment without rewriting the recording.
I see that Run Settings has "Context Parameters" - is this it?
For for a recorded .webtest, this article shows using the context parameters with the {{parameterName}} syntax for form parameters, and can indeed be used for the server name.
From inside a coded test (which inherits from WebTest), you can access it with:
this.Context["KeyName"]
Related
I have a JMeter test suite, which is working fine till apache-jmeter-2.13.
When upgrade to JMeter 3.2 and I get below exceptions in non GUI mode.
I didn't do any change in the JMeterscript, why this error is throwing in JMeter versions later 2.13 or in version 3.2?
jmeter -n -t E:\testsuite.jmx -l E:\tmp\RORlogs.csv
Error in NonGUIDriver org.apache.jorphan.util.JMeterStopTestException: ModuleCon
troller:Tracker_List_Artifacts has no selected Controller (did you rename some e
lement in the path to target controller?)
Test was shutdown as a consequence
Add an answer to close the problem:
In earlier versions of jmeter 2.13, it didn't validate the controllers which is disabled.
But from Jmeter version 3.2, irrespective of disabling all the controllers should have proper target values.
Fix the issue by assign proper target controller is assigned to disabled module controller.
Not sure if you've got the exact same situation, but what we've found is that JMeter has a problem where if you have a script with a module controller that includes another script that also has a module controller the file paths may be off and so the module controller of the included script will error.
For example. Say I have 'main.jmx' that has a module controller that points to 'external.jmx'. 'external.jmx' has a module controller that points to 'another_script.jmx'.
'main.jmx' is located in 'Documents/scripts', and this is also the directory we lauched JMeter from. 'main.jmx' has a relative path that points to 'external.jmx'. However, 'another_script.jmx' does not have the same relative path and so JMeter throws an error.
Pro tip: renaming your test elements to descriptive names will make debugging scripts far easier (it looks like you've done this, but this is more for other people reading this).
I am using Jmeter 5* and here irrespective of how many target controller names you modify, you finally need to touch all transactions/module controller from top to bottom once. Its kind of re-associating module controller with target controller. After that its working fine for me.
I used Microsoft Test Manager to create a test for a log on page (not locally hosted) and recorded this. All steps were executed without errors.
Then I've created a Coded UI test project in Visual Studio 2013.
I added a Coded UI Test with the existing recording.
I ran the test and I got the message below:
"FailedToPerformActionOnHiddenControlException: Cannot perform 'SetProperty of Text with value " on the username field.
I received this message when using the existing recording, but also if I record using the recording function in Visual Studio.
Has someone got experience with Coded UI Tests and maybe give me an example as to how to set it up.
If I understand correctly, you're basically trying to have Coded UI input a value into an HtmlInput control that's a text box on a web page. The error message "FailedToPerformActionOnHiddenControlException: Cannot perform 'SetProperty of Text with value" normally means that the value you're trying to submit isn't able to be taken by the control, either because a value isn't accepted (example, numeric values in a name field) or because it's more characters than can be accepted by the field. It could also be that the recording isn't defining the object properly, and that the property Text isn't an option on the control that was found.
So, confirm these three things:
1. The control you're tying to send value to is actually an HtmlInput. You can't send the Text property to, for example, an `HtmlDiv`.
2. The value you're submitting is a valid value for your control (you're not trying to submit numbers, for example, into a field that won't accept them).
3. The value you're submitting is within the length limit for the field.
If I were a betting man, I would say that you probably don't have the object defined properly. Take a look at the SearchProperties of the control in question first and make sure that it matches the HTML on the page itself.
Do I need to set a standard environment for testing on an other server.
I develop the Coded UI test project in a workspace on a development environment, but I need the tests to be run on a acceptation environment (the recorded tests are executed on the acceptation environment).
I read it's easy to setup this environment though, but what is best practise?
Is it better to execute it on the same environment?
Here's a code snippet:
[TestMethod]
public void TestLogonToAccount()
{
// To generate code for this test, select "Generate Code for Coded UI Test" from the shortcut menu and select one of the menu items.
this.UIMap.Enterusername();
this.UIMap.Enterpassword();
this.UIMap.Clickonlogin();
this.UIMap.ClickonCentral();
this.UIMap.Searchforemailaddress();
this.UIMap.Clickonlogin1();
}
I forgot an important part regarding the login recording test in MTM.
I just tried to play the steps again and got this error:
Playback of the selected sections of the action recording could not be completed The playback failed to find the control with the
given search properties. Additional Details: TechnologyName: 'MSAA'
Name: '' ClassName: 'MozillaWindowClass' ControlType: 'Window'
I'm working on a project regarding website test automation and I hope someone can help me with this question?
How would you recommend setting up some automated test processes that would not constantly need to be updated to test each of the core flows to test the following for a website:
login
register
sign in with facebook
save an item
delete an item
test that the few key pages (both logged in and logged out) are working like
Thanks in advance.
I'm not sure if I understand well.
But here is an instance of project setup.
Split your KeywordTest section into two folders:
A Test folder: it should contain all the KeywordTests that are called when you run your testsuite. Each of those KeywordTests should test a specific feature (Verify_Login_Fail, Verify_Login_Success...)
A Library folder: it should contain all your KeywordTests that are reusable and that might be used often from your Test folder KeywordTests. It's a kind of function library. It avoids code repetition and is easier to maintain.
For instance, you can create a KeywordTest that takes as parameter a Login and Password and that proceeds with the actions on your website to log a user in.
Store the sensible data (that might change often) in file or a database rather than hardcode it. For instance a file Login.csv where you store all the combination of login and password you want to test.
You have to write all the steps into in a class which should be in your library package and then you have to call all the methods from your test class of test package. You should use testng in your test class then create a testsuit of test class and run the script.
I'm absolutely new to CasperJS and I'm wondering what's the difference between those 'two modes'.
Both access the DOM, it seems that test mode has a limited access and functionality.
I looked for this question around and didn't find the answer.
In test mode you have access to the tester module and with it access to asserts, test suites and (xml) reports. This is not accessible in plain mode anymore (earlier versions than 1.1-beta4 had access to some of the test mode stuff in plain mode).
The only drawback to test mode is that you can only have one casper instance which is injected. This leads to:
When you try to create it, you will get an error.
(Nearly) all options have to be assigned directly and cannot be passed to create as an object.
Some things cannot be done like this one: A: How to open a new tab in CasperJS
I have had some experience writing container-bound scripts, but am totally new to web apps.
How do I debug (e.g. look at variable values, step through code etc) a web app? In a container bound script it was easy, because I could set breakpoints, use the apps script debugger - how do I go about this in a web page e.g. when I execute a doPost?
In his excellent book "Google Script", James Ferreira advocates setting up your own development environment with three browser windows; one for the code, one for the live view (in Publish, Deploy as web app, you are provided with a "latest code" link that will update the live view to the latest save when it is refreshed), and one for a spreadsheet that logs errors (using try/catch wrapped around bits of code you want to keep an eye on).
In Web Apps, even the most basic debugging of variables through Logger.log() does not work!
A great solution to have at least simple variable logging available is Peter Herrmann's BetterLog for Apps Script. It allows you to log into a spreadsheet (the same as your working spreadsheet or a separate one).
Installation is very simple - just add an external resource (see the Github readme) and a single line of code to override the standard Logger object:
Logger = BetterLog.useSpreadsheet('your-spreadsheet-key-goes-here');
Remember, that the spreedsheet that you give here as a parameter will be used for the logging output and thus must be writable by anybody!
BetterLog will create a new sheet called "Log" in the given spreadsheet and will write each log call into a separate row of that sheet.
So, for me, I debug the front-end using inspector, I haven't found a way to step through code yet, but you can use 'debugger' in your javascript (along with console.log) to stop the code and check variables.
to debug the backend, what I've been doing is to write my functions like
function test_doSomething(){
payload = "{item1: 100, item2: 200}" //<- copy paste from log file
backend_doSomething(payload)
}
function backend_doSomething(payload){
Logger.log(payload)
params = JSON.parse(payload)
...
}
Then after refreshing your project on the backend, you can look at executions, grab the payload from the log file, and paste it into your test_doSomething() function.
From there, you are re-creating the call that you want to debug and you can run that, stepping through the backend code as usual.