Testing Service Fabric deployment packages on VSTS

The scenario – you have a Service Fabric build configured on Visual Studio Team Services (VSTS) as follows:


As you can see from the screenshot, there is a task to generate the Service Fabric deployment package. There were no errors in this task, but don’t assume that everything is OK with the package, something might go wrong when you try to deploy it to a SF cluster.

In order to avoid surprises when deploying the application, you can test the package after its generation using the Test-ServiceFabricApplicationPackage powershell cmdlet.

Add a new Powershell++ task after generating the package and configure it as follows:

The command takes the path to the SF package folder as a parameter. I usually set the SF project as the working folder.

Queuing a new build, you can see the results of the build and in particular the task that tests the SF package:


That’s it! With this solution you will know immediately if something is wrong with the package, saving you from the frustration of a failed deployment. This doesn’t mean that deployments will never fail, but hopefully you will be able to detect most or all of the errors in the deployment package every time you trigger a new build 🙂

Improving the performance of Service Fabric builds and deployments on VSTS

Consider the following definition – this is my typical build for a Service Fabric (SF) application on Visual Studio Team Services (VSTS):


In short, after restoring the nuget packages, building the solution and running the unit tests I generate the Service Fabric deployment package and test it. The artifacts of the build are not only the deployment package but also the publish profiles for each environment, as follows:


Unlike an ASP.NET application the Service Fabric deployment package is not zipped, as you can see in the above screenshot. Publishing the artifacts in this case takes around 11 seconds:


What does it mean exactly “publishing the artifacts”? It means that the files files will be uploaded to the server. Obviously the more and bigger files you have the longer it will take to upload them to a server (and later on downloading them when you need to deploy the application).

I decided to zip the deployment package instead to improve the performance. I added another task as follows:


And this is the deployment package, zipped:


Zipping and uploading the deployment package took around 8 seconds:


You might think that’s not a big improvement but times will change depending on the size of the deployment packages. In this case the generated folder has 26MB but I’ve heard of deployment packages that have almost 200MB in size!

Also, have in consideration that when you do a deployment you’ll need to download the artifacts from the server. In the first case the deployment package (uncompressed) took on average 10 seconds to download:


As opposed to less than 3 seconds in the second case (zipped deployment package):


That’s it! In this example I have shown you a Service Fabric build but I’d recommend this approach if you upload multiple files as artifacts. Even though there are extra tasks in the build (zip the package) and release (unzip the package) there was an improvement in the performance of both the build and release.

My suggestion is to create 2 versions of the same build and also the release and compare the results between each version. Give it a try, you might be surprised with the result!

Happy coding 🙂

Powershell scripts running on Bamboo don’t return the correct exit code

As part of an deployment project on Bamboo CI, I was running a powershell script to deploy an ASP.NET application to a Cloud Service on Azure.

Even though there was an error executing the script, Bamboo was setting the status of the Deployment to Success. Why? Because the exit code returned by the powershell script is always 0 (zero means successful execution).

After some research I was able to find a way to return the correct exit code in case of failure. I added the following lines to the top of my powershell script:

    write-output $_
    exit 1

The trap statement includes a list of statements to run when a terminating error occurs – in this case, every time an error occurs the error message will be displayed and then the script will return a correct exit code indicating a failure. I am returning 1 but any value different from 0 (zero) will do the trick 🙂

Running tests in Bamboo after a deployment

I’ve been using Bamboo CI Server for the last few months to automate builds and deployments. I like the tool because it has good integration with Jira (both tools are from Atlassian), it’s easy enough to configure new builds and deployments, triggers, notifications, etc.

But I realised that something important was missing: Bamboo allows you to add a test runner task in a build project but not in a deployment project! This means that you can’t run tests after a successful deployment (smoke tests, integration tests, …), at least not without a workaround.

The trick is to configure your test runner as an executable in Bamboo. These are the steps in order to configure NUnit and run tests in a deployment project (it should work for any other test runner):


1. Add a new executable for NUnit

Go to Bamboo Administration and click on “Executables” on the left panel.

01 bamboo administration

Click on “add an executable as a server capability

02 click link

Add the path to NUnit Console and a label for the new executable. It is important to set the type to “Command” in order to use it in a Deployment project:

03 add-executable

Click on the “Add” button to save the new command.


2. Add a new deployment task to run the tests

You can either add a new task for the tests to an existing deployment or add a new deployment project that will only run the tests.

I decided to add a new deployment project that will be triggered after a successful deployment because it’s easier to understand if there is actually a problem with the deployment itself or if the integration tests are failing. Also, this way I am able to run the tests at any time without having to deploy the application.

Whatever your choice is, add a new “Command” task to the deployment project:

04 - add-new-task

In the “Executable” dropdown you should be able to find the command you configured for NUnit. Add arguments and environment variables if necessary:

06 - configure-nunit-task

Save the task and run the deployment. This is an excerpt of the generated log that contains the test results:

NUnit-Console version
Copyright (C) 2002-2012 Charlie Poole.
Copyright (C) 2002-2004 James W. Newkirk, Michael C. Two, Alexei A. Vorontsov.
Copyright (C) 2000-2002 Philip Craig.
All Rights Reserved.

Runtime Environment - 
   OS Version: Microsoft Windows NT 6.2.9200.0
  CLR Version: 2.0.50727.8009 ( Net 3.5 )

ProcessModel: Default    DomainUsage: Default
Execution Runtime: net-4.0
Tests run: 5, Errors: 0, Failures: 4, Inconclusive: 0, Time: 6.8491962 seconds
  Not run: 0, Invalid: 0, Ignored: 0, Skipped: 0

Errors and Failures:
1) Test Failure : GivenAnUrl_WhenGettingPage_ShouldreturnSuccessStatusCode("/Home.aspx")
     Expected: True
  But was:  False

2) Test Failure : GivenAnUrl_WhenGettingPage_ShouldreturnSuccessStatusCode("/Services/Activate.aspx")
     Expected: True
  But was:  False

3) Test Failure : GivenAnUrl_WhenGettingPage_ShouldreturnSuccessStatusCode("/Administration/LostPassword.aspx")
     Expected: True
  But was:  False

4) Test Failure : GivenAnUrl_WhenGettingPage_ShouldreturnSuccessStatusCode("/Shop/Product/List.aspx")
     Expected: True
  But was:  False

Failing task since return code of [C:\Program Files (x86)\NUnit 2.6.4\bin\nunit-console.exe integration-tests-uat.nunit --config="release"] was 4 while expected 0
Finished task 'Run integration tests' with result: Failed
Finalising the build...
Stopping timer.
Build 12484609-16973828-16613398 completed.
Finished processing deployment result Deployment of 'release-16' on 'UAT - Integration Tests'

That’s it! The output is not nicely formatted as in the build tasks but it does the job – you can see how many tests were run and how many have failed (if any).

Managing deployments to Azure

 The scenario – I am working on an ASP.NET web site that is hosted on Azure as a Cloud Service.

I have automated both the build and the deployment to the cloud using the Bamboo build server. The build compiles the solution, runs the unit tests and generates the deployment packages that are used by Bamboo to deploy the site to Azure.

I could use Bamboo to manage all deployments but given that Cloud Services offer out-of-the-box support for blue-green deployments I decided to take advantage of both systems. This is how I have organised things:

1. Deployment to Staging

Deployment to Staging is performed by Bamboo. Automated tests are executed as part of the deployment to check if everything is working as expected. Additional manual tests should be executed as well before deploying to Production.


2. Deployment to Production

This is done using the Azure Management Portal – simply click on the Swap button and traffic will be routed to the Staging environment, which now becomes the Production environment.


3. Rolling Back

Something went wrong? No problem, click on the Swap button again to switch the environments – it’s just as simple as that.


Final thoughts

Azure Web Sites or Cloud Services offer out-of-the-box support for blue-green deployments, which provide a simple and powerful way to test a deployment before going to production and roll it back, if necessary.

You can use your Continuous Delivery server in conjunction with Azure Management Portal to manage your deployments to the cloud – consider all the advantages and disadvantages and use the functionalities of each system that makes your life easier 🙂