I have a test plan(Contains test cases linked to UI tests. hence automated). To run them, i have linked them to a release definition(Continuous deployment). All tests pass, but not UI.
I understood that it does not have an .exe to run. Hence, i've made sure that necessary files + .exe from project's bin folder gets copied to UI project's bin folder.
But, when I see the output in server while creating release, only, dll of test project was copied not .exe files.
I need help in writing my release definition to run all test plans and deployment. I've used TFS 2017, vsTest(version 2), VS 2017
Related
My ASP .NET Solution (net framework 4.7.2) consists of several projects. I am interested in having a build definition (Azure) that will run my tests from a specific folder ( that folder contains some tests projects -nunit tests )
I have searched for information about this, but most solutions seemed constructed towards building & releasing, with all the tests in the whole solution being run.
How should I approach this? I will need to build my solution (Visual studio build ) as a task (?), but what should be the steps(tasks) to select and run my test folder ?
You could actually do something like this How to run a list of tests with VsTest.Console on windows agents if you prepare such file with list of tests from that folder.
You can find vstest.console.exe on this location on host agents
C:\Program Files (x86)\Microsoft Visual Studio\2019\Enterprise\Common7\IDE\Extensions\TestPlatform\vstest.console.exe
In that case you just need to prepare such file, put it in source control and run in your pipeline.
I will need to build my solution (Visual studio build ) as a task (?)
The answer is yes. That because the folder **\bin\** is ignored by source control by default. We need to build the solution in our build pipeline to generate dll files.
but what should be the steps(tasks) to select and run my test folder ?
We could specify the specific folder in the Visual Studio Test task in the Search folder:
Or specify that folder in the option Test files:
I have a VS2010 solution that includes a custom project. The project builds an installer out of all the libraries and executables built in the same solution. This is a third-party installer builder, not a native Visual Studio installer project. It runs a custom command in the post-build event of the project.
The problem with this setup is that the installer project build runs (and succeeds!) even if some other project build fails.
I have added project dependencies so that the installer project depends on all other projects, to no avail. I have also tried to add project references with the same result.
This happens only with this specific project (probably because it's a custom project with configuration type "utility"). Other projects don't start building if a dependent project fails to build.
This is a dangerous situation. A developer may not pay attention to a build failure in some other project and use the incorrectly built installer. I need MSVC to skip building the installer project if some other project fails.
How should I deal with it?
Ideally this should be solved at the MSVC level. I don't want to add custom checks to the installer build command because this would mean I have to maintain a list of projects/targets in two separate places. I also don't want to introduce additional tools to the picture.
I am have built an MSBuild project that has
the main .proj file with several Targets inside,
several .targets files
and several .rsp response files to run this project in several ways. Such as
Build, Clean/Build/Deploy, Clean/Build/Test/Deploy.. You get the idea.
What I now want to do is take this MSBuild and run it in TFS scheduler after I pull the source code. So the workflow should be
1. Pull the source code in TFS
2. Run MSBuild project in the scheduler so I might set up tasks to run hourly and nightly.
The MSBuild will take care of Deploying to IIS, unless someone has a more efficient way of deploying after an hourly build.
How can I accomplish this in TFS?
Thank you
You can use the Team Explorer UI in Visual Studio to define a "scheduled" build definition.
http://visualstudiomagazine.com/articles/2012/04/11/creating-a-build-definition-in-tfs.aspx
If you want to take this to the next level, research "continuous integration" which is also a built in capability.
I recently finished a solution composed by 2 project. For this 2 project I have the .exe build, but I don't know how to do the exe of solution with VS 2012. Someone know how to do it?
When you compile & build a solution, the .exe that is created is placed into the folder relating to your current build configuration, usually either Debug or Release.
So navigate to the folder, on disk, using Windows Explorer, where the solution is stored, and look in the Debug folder. The compiled program is there.
If this is not what you are after, please post more details about these two projects and how they 'tie together'.
So I've downloaded VS2012 RTM and upgraded my Coded UI test projects to the VS2012 versions.
The reason I have done this is because I was having issues with VS2010 Coded UI tools due to partial support of IE9.
So I ran the tests from my developer machine both locally and remotely (making use of the new Visual Studio agent tools), and the tests worked successfully. Presumably, the reason my tests now work is that VS2012 upgraded them so that they now work with IE9, right?
So - this is my problem. I want to automate these tests, by launching them from our build machine - but our build machine still runs on VS2010, and for the time-being this is not going to change.
So I checked in my new VS2012 Coded UI test solution into TFS and queued a new build - so the build machine built my solution. And the build was successful. All good there.
So, next I created a new test case in Microsoft Test Manager and associated with the ordered test list in my new solution. Then I launched the tests (using the existing VS2010 test controller) on my remote test environment (which has the existing VS2010 test agent tools).
But the tests failed - the same issues that affected my tests with the VS2010 Coded UI tests (due to lack of full support for IE9)
Why did they fail?
Do I need the new VS 2012 agent tools for my test controller and agents? Must I build my solution in VS2012?
Ideally, I don't want to have to install VS2012 RTM on my build machine - I want to do the minimum possible to get my tests working and automating.
Is there a way around this?
Your coded ui tests reference assemblies that come with VS installation or your agent's installation (the assemblies defininf the WpfControl, Mouse, Keyboard, Playback and other classes).
So if you run it on a build/test machine with older versions of the dll's the same issues are going to stay, the tests will not start using the newer fixed assemblies thatt were shipped with VS2012.
As a temporary workaround you can check what dll's you reference and make sure that the build process puts them in the same directory with your tests assembly. That way the lookup for them will find the in the current directory without using the /path and finding them in the VS installation directory.