On September 10, 2018, Microsoft announced Azure DevOps, which is the next phase of software delivery. Azure DevOps is the next evolution of Visual Studio Team Services (VSTS). Quick side note, VSTS users will be upgraded automatically to the new Azure DevOps structure. It will help organizations delivery faster and with high quality across a wide breath of the development life-cycle and systems.
Let’s see the official announcement from Donovan Brown the DevOps Manager at Microsoft.
Azure DevOps will include the following:
- Azure Boards – a powerful way to track deliverable through boards, backlog, dashboards, and reporting.
- Azure Repos – an unlimited cloud hosted repositories for source control and collaboration.
- Azure Artifacts – Various development package feeds from public and private sources like npm and NuGet.
- Azure Pipelines – CI\CD that can deliver to any language, platform, or cloud.
- Azure Test Plans – Testing, Testing, Testing – all in one solution.
Each Azure DevOps service is open and extensible and works great for any type of application regardless of the framework, platform, or cloud. Learn more about cost from Azure DevOps Pricing.
Here is Edward Thomson, the Program Manager for Azure DevOps discussing Azure Pipelines, which is on of the key pillars in Azure DevOps.
There are some key changes to this transition of VSTS To Azure DevOps that you will see:
- URLs will change from <name>.visualstudio.com to dev.azure.com/<name>.
- Services will have an updated user experience.
- On-premise TFS will receives updates based on new features and will renamed to Azure DevOps Server.
You might want to visit the save-the-date and watch live streams on Azure DevOps events page. There you’ll also find additional on-demand videos and other resources to help get you started using Azure DevOps.
Recently I inherited some poorly written WebJobs that I needed to create an automated deploy for a client to their various environments (DEV, QAE, and PRD). To start the WebJobs were missing the publish package required to publish them to Azure, which is added either by the package manager (seen bleow) or the Publishing via Visual Studio (using 2017) Publish.
You can add the publishing package from NuGet by running the following command in Package Manager console:
Once that was completed, we need to head over the Visual Studio Team Services (VSTS) to create a build and release for these WebJobs. The tests were disabled as there were not any in the project and planned to add them in the future.
You only need to do a few modifications to canned build template; however, I did add some build arguments as follows:
/p:DeployOnBuild=true /p:WebPublishMethod=Package /p:PackageAsSingleFile=false /p:SkipInvalidConfigurations=true /p:PackageLocation="$(build.artifactstagingdirectory)\\Webjob Deploy" /p:PackageTempRootDir="\Web"
One of these arguments changes the deployment path to “Webjob Deploy” and in the Publish Build Artifacts task I change the Path to Publish to the following:
Once I ran the build my artifacts looked like:
Once this is complete we can head over to my build the release for the WebJob code. We start by linking the artifacts and create an environment, I general start with one to get it working and tested then clone it and change the variables for that environment.
Part of the release was that I needed to create the Azure Resource Group and deploying all the components then linking my deployment artifacts to my these resources (App Service and App Service Plan). I plan to add a blog post on DSC (Desired State Configuration) scripts and ARM (Azure Resource Manager) templates, so more to come.
Hope this journey helps!
Important information I gathered from Logic Apps on the documentation found at https://azure.microsoft.com/en-us/services/logic-apps/ and will be adding a Logic App walk-through.
What are Logic Apps?
- Part of iPaaS (integration Platform as a Service), which allow them to scale automatically on demand
- Help simplify scalable integrations and workflow in the cloud, like BizTalk server does for on premise solutions.
- Use visual workflows and connectors to create integrations
- Can be used with custom APIs, code, and actions
- Provides Managed Connectors for fast and consistent integration.
Connectors supported – https://azure.microsoft.com/en-us/documentation/articles/apis-list/
Logic App Uses
- Provide a variety of triggers and actions with management tools to help centralize your API development.
- Workflows – graphical way to model your business processes as a series of steps or a workflow.
- Managed Connectors – Managed connectors are created specifically to aid you when you are connecting to and working with your data.
- Triggers – Some Managed Connectors can also act as a trigger. A trigger starts a new instance of a workflow based on a specific event, like the arrival of an e-mail or a change in your Azure Storage account.
- Actions – Each step after the trigger in a workflow is called an action. Each action typically maps to an operation on your managed connector or custom API apps.
- Enterprise Integration Pack – For more advanced integration scenarios, Logic Apps includes capabilities from BizTalk. BizTalk is Microsoft’s industry leading integration platform. The Enterprise Integration Pack connectors allow you to easily include validation, transformation, and more in to your Logic App workflows.
- Single Http request and/or Connector call limits
- Request Timeout is 1 minute
- Message size is limit is 50mb, with request trigger support for up to 25MB
- Expression evaluation is limited to 131,072 character meaning @cancat(), @base64(), string cannot be longer
- Retry policy is max delay is 1 hour with a min of 20 minutes with a max of 4 retry attempts
- Run duration and retention is limited to a run duration of 90 day with a storage retention of 90 days with min recurrence intervals of 15 sec and max recurrence interval of 500 days.
- Looping and debatching limits for ForEach items, SplitOn item, Until iterations to 10,000 and ForEach Parallelism to 20.
- Throughput limits for a single app instance are 100 triggers per second.
- Definition limits Actions in ForEach to 1 with Action per workflow to 60 and action nesting depth to 5. Triggers per workflow are limited to 10 with a max character per expression to 8192. Action/Trigger names are limited to 80 characters with a description length to 256. Parameters are limited to 50 and outputs are limited to 10.
Three different trigger statuses:
- Skipped. It polled the endpoint to check for data and received a response that no data was available.
- Succeeded. The trigger received a response that data was available. This could be from a manual trigger, a recurrence trigger, or a polling trigger. This likely will be accompanied with a status of Fired.
- Failed. An error was generated.
A run displays one of the following statuses:
- Succeeded. All actions succeeded, or, if there was a failure, it was handled by an action that occurred later in the workflow. That is, it was handled by an action that was set to run after a failed action.
- Failed. At least one action had a failure that was not handled by an action later in the workflow.
- Cancelled. The workflow was running but received a cancel request.
- Running. The workflow is currently running.
While updating an ASP.NET MVC website running Angular I ran across a $Modal issue and the error was Unknown provider: $modalProvider <- $modal. In order to fix this issues I had to change the $Modal inject to $uiModal, which was due to a naming change in the angular-ui-bootstrap framework.
The framework is located at: https://angular-ui.github.io/bootstrap
I had to make some updates in one of to a clients Azure SQL databases, that required me to drop and recreate a table that had a foreign key relationship. To do this I needed to drop and recreate the FK relationship on a table and I used this script to find all constraints in the the database.
FK_Table = FK.TABLE_NAME,
FK_Column = CU.COLUMN_NAME,
PK_Table = PK.TABLE_NAME,
PK_Column = PT.COLUMN_NAME,
Constraint_Name = C.CONSTRAINT_NAME
INNER JOIN INFORMATION_SCHEMA.TABLE_CONSTRAINTS FK
ON C.CONSTRAINT_NAME = FK.CONSTRAINT_NAME
INNER JOIN INFORMATION_SCHEMA.TABLE_CONSTRAINTS PK
ON C.UNIQUE_CONSTRAINT_NAME = PK.CONSTRAINT_NAME
INNER JOIN INFORMATION_SCHEMA.KEY_COLUMN_USAGE CU
ON C.CONSTRAINT_NAME = CU.CONSTRAINT_NAME
INNER JOIN (
INNER JOIN INFORMATION_SCHEMA.KEY_COLUMN_USAGE i2
ON i1.CONSTRAINT_NAME = i2.CONSTRAINT_NAME
i1.CONSTRAINT_TYPE = 'PRIMARY KEY'
ON PT.TABLE_NAME = PK.TABLE_NAME
Unfortunately you are not able to run both side by side; so I would suggest you run 2009 in a VM and test your 2010 project in that if you need downward compatibility.
Remember, BizTalk 2010 is the only version of BizTalk compatible with R2 version of the Microsoft Server 2008 and SQL 2008. Setup is still roughly the same.
I have not been able to find a post or get a true explanation on “reserved words” in BizTalk; however there are a few you should stay away from. I have personally tested these.
You should refrain from using “BTS” or “EDI” anywhere in your namespace to avoid namespace property look-up issues.
I will be posting more about this as I work through the public version of BizTalk 2010 on Oct. 1, 2010.
Public production date announced was Oct. 1, 2010.
Free Developer Version Available