We tried to implement the dependency mapping explained in this Microsoft post, but it doesn't work.
Microsoft dependency mapping implementation example
We have enabled in our build the "Run only impacted test" and set the "Settings file" and TIA.usermapfile from visual studio task.
After that we have created the XML file at the root of the solution, and set parameters for executing specific test from FullyQualifiedName contained in the changes from check-in.
We have verified that the XML file is correctly located at the $(System.DefaultWorkingDirectory) in the build agent server.
We have checked in some modifications which impact the repositories classes.
As you can see, there is no information about the TIAmap.xml file in the Test assemblies log, and the tests were not started, whereas the code changes impact the repository folder.
We have more two thousand tests and there are none that are launched.
Do you have any idea why the tests that would impact and not impact the repositories folder aren't run?
Related
We have a build definition for our project that's set to continuous integration and which runs all the unit tests on every check in. It was all running fine until recently, when it began to reject a handful of tests that worked on developer machines.
A moment's examination revealed that this was because the tests were dependent on an external file template for which the path was specified in the app.config file. It was a network drive which developer machines had access to, but which didn't exist on the build server.
The project is a WPF application. Ideally I'm going to try and refactor the code to see if I can bypass that external dependency. But if I can't, or just to satisfy my curiosity, is there a way of editing the build itself to use or deploy a different config file?
You have a much bigger problem: Your code and your tests are dependent on some random, external, uncontrolled file. That file should be something that's source controlled and deployed along with your application.
You also have an external dependency in your tests: It has to read that file from the file system. You may want to break that dependency using standard inversion of control techniques so that tests can run in true isolation, without any need for a file to be present in the file system.
SlowCheetah should work for this use case.
Nuget.org - SlowCheetah
I have recently added a TFS build to our project, and configured it to run the automated tests that are in the project, and for some reason the build is ignoring the tests as if they don't exist!
In order to figure this out I created another solution with only a basic project from the origin code and the test project as well.
I've added these to the TFS and configured an almost equal build to this solution, and guess what? it executed the tests! it's the same tests as the original. just a copy of it.
The main difference between these 2 solutions is that the original code is a big solution with many projects, most of the projects are in some kind of a solution directory (and so does the tests project - it's inside a solution directory as well).
The difference between the TFS build definition is only that the output location of the build is set to single directory (I also tried PerProject and it worked) and in the original code it's defined to "AsConfigured" because we have some build tasks for copying dlls and such.
Has anyone encountered this problem?
Any ideas?
Thanks
TFS will look for the test DLLs in the BINARIES output folder location. You need to configure your 'build tasks for copying DLLs' to ensure your test DLLs are copied to this location.
Test assembly file specification
Specify the binary files that contain the tests that you want to run. Leave the default value (**\**test*.dll) if you want the build agent to search recursively for any .dll files that match *test*.dll in the binaries subdirectory of the build agent's working directory.
TFS2013 — http://msdn.microsoft.com/en-us/library/ms253138.aspx
TFS2012 — http://msdn.microsoft.com/en-us/library/ms253138(v=vs.110).aspx
I'm just looking at streamlining the nuget packages that are coming out of my build system and I'm stuck on how to only package the files that are required.
I have several configurations sharing a Root VCS checkout. I have a configuration that runs a debug build with unit tests. I also have a release configuration that does a release build, this configuration then also uses the TeamCity OctoPack plugin to create the nuget packages.
What I want to achieve is the building of nuget packages that don't contain the *.pdb and *.xml documentation files as these aren't required for the release deployment.
I've looked through this page on the OD site:
http://docs.octopusdeploy.com/display/OD/Using+OctoPack
And according to this page OctoPack should only package up the required files by default. I'm not entirely clear on how or what needs to be done to get around this problem as it doesn't appear to be working as described.
It seems that one solution would be to provide a nuspec file for the projects I'm looking to deply but I'm also wondering if there is something I'm missing before I head off down that route.
I also have some MEF plugins that are copied in post build events and these aren't included in the nuget packages when in fact they are needed for the application to run. I think I need to get explicit with a nuspec file but would like to confirm this.
What is the simplest way of achieving what I need?
Assuming you're running the later versions of OctoPack, in your release build you can set a system parameter system.DebugType = None which will get passed to the OctoPack build scripts and prevent the PDB's being created.
This simply overrides the setting defined in your csproj msbuild file (assuming C#), so you can use it wherever you want to prevent PDB's being created at the build configuration level (not just OctoPack). I generally prefer this approach as it prevents side-effects in your build from changes by developers in the project file.
As for the xml files, I haven't actually tried this, but you can try a similar approach and create a system parameter system.DocumentationFile = "" to blank out the output.
If you really want to make sure that the files have been removed there are a couple of ways you can do this. Modify your deployment process to:
Execute your own custom PowerShell script in that removes the files
Include a script module from the Octopus Library to the same. Check out the File System - Clean Directory from the Octopus Library
We are using TFS 2010. In our projects we have multiple configuration files. Each configuration file is needed for a specific build.
Example: Project XYZ includes three configuration files, lets say DEV-CONFING, QA-CONFIG and PROD-CONFIG files.
Now we want to find a way that during the build process, the build process includes the right configuration file and excludes the rest.
Example:
Build for Dev should pickup and copy only the DEV-CONFIG file (and excludes all other configuration files)
Build for QA should pickup and copy only the QA-CONFIG file (and excludes all other configuration files)
Build for production should pickup and copy only the PROD-CONFIG file (and excludes all other configuration files)
Is it possible to change the setting of individual build definition and address this issue?
What is best possible solution for this problem in VS 2010 and TFS 2010?
Please list all required steps and be precise as much as possible because I'm a newbie in TFS 2010 and have not much experience with the whole process.
The "Web Config Transformations in VS2010" addresses the problem of modifying the web.config during compile time. Some people might want the web.config to be modified at deployment time instead of compile time (like me) so that we dont have to recompile the code while moving from one environment to another, in that case you can make use of "Web Deployment Parameters"
You can read more about it at:
http://www.asp.net/mvc/tutorials/deployment/visual-studio-web-deployment/web-config-transformations
web config transformation syntax:
http://msdn.microsoft.com/en-us/library/dd465326.aspx
#user2585405 to your comment above "But all the configurations in my case are for the same mode "Release Mode". The decision should relay on the build definition. So when I use a build definition for DEV, the configuration for DEV should be automatically chosen or for QA build definition, the corresponding QA configuration should be automatically chosen."
Doesnt matter whether currently you have only one configuration. You can add how much ever configurations you want, right click the solution and select the configuration manager. Now add a new configuration say "Dev" (it can be in "Release/Debug" and also select the platform "x86, any cpu etc"). Now add another configuration for QA and PROD. Then come back to your web.config and right click and select "Add web config transformation".
Now you should be able to see the web.Dev.Config, web.QA.config etc. The root web.config will have all the common properties and the DEV/QA/Prod specific config will have only the delta changes (you have to define these entries) which are specific to that configuration (lets say connectionstring). When you compile, within the build definition you can select which configuration you want to compile (Dev, QA, prod) you dont have to keep multiple build defintion. Just one is enough! After the compilation is done, the root web.config will be replaced with the values specific to the configuration for the connectionstring part alone.
Hope this helps!
Yes this possible. You can keep configuration files in different mode for eg:
You can have configuration files in Release mode debug mode, release mode and the common configuration file used by all.
So when you build your solution in Debug mode then the debug mode configuration file will be used and when in release mode then release mode configuration file will be used.
For further clarity take example of web config when you will expand the web config tab you will see the two configuration file in different modes.
I am setting up a Continuous Integration server.
I have one issue that doesn't seem to be mentioned in the tutorials.
I have a ASP.net Web Application that I need to compile and them publish.
My Problem is that I seem to be able to compile the app but when I attempt to use a buildPublisher this copies every thing including .svn files & folders and ms CS files.
I am using an MSBuild task to compile my source. I tried setting my MSBuild Output Directory to directory but this didn't seem to have any effect.
What am I not understanding?
Thanks
You're probably looking for the _CopyWebApplication directive:
http://blogs.msdn.com/nmoreau/archive/2007/01/26/deploying-web-application-projects.aspx
We use CruiseControl.NET to deploy our ASP.NET applications to our test servers for the QA department so this is indeed possible.
In each project we created an additional build configuration called 'Deployer' which is identical to the Debug configuration with the exception of building an additional Web Deployment Project. So by running MSBuild in the Deployer configuration we can generate our compiled output in a known location.
We then use NAnt to perform a simple copy operation to the required location, i.e. into a folder where our IIS server is configured to look for the applications.
I know you don't mention NAnt in your question but it's well worth getting familiar with it if you want to get the most out of CCNet.
I'm not at work at the moment but if this makes any sense and you want some additional information then let me know and I'll pull some more information together.
Hope this helps
Are your bin or obj folders checked into svn? if so that would contribute or possibly cause this issue.
Because those folders shouldn't have any svn folders/files
You can configure SVN to call the .svn folders _svn instead.
Set the SVN_ASP_DOT_NET_HACK environment variable.