Dialogporten

Getting started with local development

Mac

Prerequisites

• .NET 8 SDK (see global.json for the currently required version)

Installing Podman (Mac)

1. Install Podman

2. Install dependencies:

brew tap cfergeau/crc
# https://github.com/containers/podman/issues/21064
brew install vfkit
brew install docker-compose

3. Restart your Mac

4. Finish setup in Podman Desktop

5. Check that Docker Compatility mode is enabled, see the bottom left corner

6. Enable privileged testcontainers-dotnet
   echo "ryuk.container.privileged = true" >> $HOME/.testcontainers.properties

Windows

Prerequisites

• Git
• .NET 8 SDK
• WSL2 (To install, open a PowerShell admin window and run wsl --install)
• Virtual Machine Platform (Installs with WSL2, see the link above)

Installing Podman (Windows)

1. Install Podman Desktop. .

2. Start Podman Desktop and follow instructions to install Podman.

3. Follow instructions in Podman Desktop to create and start a Podman machine.

4. In Podman Desktop, go to Settings → Resources and run setup for the Compose Extension. This will install docker-compose.

Running the project

You can run the entire project locally using podman compose. (This uses docker-compose behind the scenes.)

podman compose up

The following GUI services should now be available:

• WebAPI/SwaggerUI: localhost:7124/swagger
• GraphQl/BananaCakePop: localhost:7215/graphql
• Redis/Insight: localhost:7216

The WebAPI and GraphQl services are behind a nginx proxy, and you can change the number of replicas by setting the scale property in the docker-compose.yml file.

Running the WebApi/GraphQl in an IDE

If you need do debug the WebApi/GraphQl projects in an IDE, you can alternatively run podman compose without the WebAPI/GraphQl.
First, create a dotnet user secret for the DB connection string.

dotnet user-secrets set -p .\src\Digdir.Domain.Dialogporten.WebApi\ "Infrastructure:DialogDbConnectionString" "Server=localhost;Port=5432;Database=Dialogporten;User ID=postgres;Password=supersecret;Include Error Detail=True;"

Then run podman compose without the WebAPI/GraphQl projects.

podman compose -f docker-compose-no-webapi.yml up 

DB development

This project uses Entity Framework core to manage DB migrations. DB development can either be done through Visual Studios Package Manager Console (PMC) or through the CLI.

DB development through PMC

Set Digdir.Domain.Dialogporten.Infrastructure as the startup project in Visual Studio's solution explorer, and as the default project in PMC. You are now ready to use EF core tools through PMC. Run the following command for more information:

Get-Help about_EntityFrameworkCore

DB development through CLI

Install the CLI tool with the following command:

dotnet tool install --global dotnet-ef

You are now ready to use EF core tools through CLI. Run the following command for more information:

dotnet ef --help

Remember to target Digdir.Domain.Dialogporten.Infrastructure project when running the CLI commands. Either target it through the command using the -p option, i.e.

dotnet ef migrations add -p .\src\Digdir.Domain.Dialogporten.Infrastructure\ TestMigration

Or change your directory to the infrastructure project and then run the command.

cd .\src\Digdir.Domain.Dialogporten.Infrastructure\
dotnet ef migrations add TestMigration

Testing

Besides ordinary unit and integration tests, there are test suites for both functional and non-functional end-to-end tests implemented with K6.

See tests/k6/README.md for more information.

Updating the SDK in global.json

When RenovateBot updates global.json or base image versions in Dockerfiles, make sure they match. The global.json file should always have the same SDK version as the base image in the Dockerfiles. This is to ensure that the SDK version used in the local development environment matches the SDK version used in the CI/CD pipeline. global.json is used when building the solution in CI/CD.

Development in local and test environments

To generate test tokens, see https://github.com/Altinn/AltinnTestTools. There is a request in the Postman collection for this.

Local development settings

We are able to toggle some external resources in local development. This is done through the appsettings.Development.json file. The following settings are available:

"LocalDevelopment": {
    "UseLocalDevelopmentUser": true,
    "UseLocalDevelopmentResourceRegister": true,
    "UseLocalDevelopmentOrganizationRegister": true,
    "UseLocalDevelopmentNameRegister": true,
    "UseLocalDevelopmentAltinnAuthorization": true,
    "UseLocalDevelopmentCloudEventBus": true,
    "UseLocalDevelopmentCompactJwsGenerator": true,
    "DisableCache": true,
    "DisableAuth": true,
    "UseInMemoryServiceBusTransport": true
}

Toggling these flags will enable/disable the external resources. The DisableAuth flag, for example, will disable authentication in the WebAPI project. This is useful when debugging the WebAPI project in an IDE. These settings will only be respected in the Development environment.

Using appsettings.local.json

During local development, it is natural to tweak configurations. Some of these configurations are meant to be shared through git, such as the endpoint for a new integration that may be used during local development. Other configurations are only meant for a specific debug session or a developer's personal preferences, which should not be shared through git, such as lowering the log level below warning.

The configuration in the appsettings.local.json file takes precedence over all other configurations and is only loaded in the Development environment. Additionally, it is ignored by git through the .gitignore file.

If developers need to add configuration that should be shared, they should use appsettings.Development.json. If the configuration is not meant to be shared, they can create an appsettings.local.json file to override the desired settings.

Here is an example of enabling debug logging only locally:

// appsettings.local.json
{
    "Serilog": {
        "WriteTo": [
            {
                "Name": "Console",
                "Args": {
                    "outputTemplate": "[{Timestamp:HH:mm:ss.fff} {Level:u3}] {Message:lj}{NewLine}{Exception}"
                }
            }
        ],
        "MinimumLevel": {
            "Default": "Debug"
        }
    }
}

Adding appsettings.local.json to new projects

Add the following to the Program.cs file to load the appsettings.local.json file:

var builder = WebApplication.CreateBuilder(args);
// or var builder = CoconaApp.CreateBuilder(args);
// or var builder = Host.CreateApplicationBuilder(args);
// or some other builder implementing IHostApplicationBuilder

// Left out for brevity
builder.Configuration
    // Add local configuration as the last configuration source to override other configurations
    //.AddSomeOtherConfiguration()
    .AddLocalConfiguration(builder.Environment);

// Left out for brevity

Pull requests

For pull requests, the title must follow Conventional Commits. The title of the PR will be used as the commit message when squashing/merging the pull request, and the body of the PR will be used as the description.

This title will be used to generate the changelog (using Release Please) Using fix will add to "Bug Fixes", feat will add to "Features". All the others,chore, ci, etc., will be ignored. (Example release)

Deployment

This repository contains code for both infrastructure and applications. Configurations for infrastructure are located in .azure/infrastructure. Application configuration is in .azure/applications.

Deployment process

Deployments are done using GitHub Actions with the following steps:

1. Create and Merge Pull Request

• Action: Create a pull request.
• Merge: Once the pull request is reviewed and approved, merge it into the main branch.

2. Build and Deploy to Test

• Trigger: Merging the pull request into main.
• Action: The code is built and deployed to the test environment.
• Tag: The deployment is tagged with <version>-<git-sha>.

3. Prepare Release for Staging

• Passive: Release-please creates or updates a release pull request.
• Purpose: This generates a changelog and bumps the version number.
• Merge: Merge the release pull request into the main branch.

4. Deploy to Staging (Bump Version and Create Tag)

• Trigger: Merging the release pull request.
• Action:
  • Bumps the version number.
  • Generates the release and changelog.
  • Deployment is tagged with the new <version> without <git-sha>
  • The new version is built and deployed to the staging environment.

5. Prepare deployment to Production

• Action: Perform a dry run towards the production environment to ensure the deployment can proceed without issues.

6. Deploy to Production

• Trigger: Approval of the dry run.
• Action: The new version is built and deployed to the production environment.

Visual Workflow

Release Please is used to create releases, generate changelog and bumping version numbers.

CHANGELOG.md and version.txt are automatically updated and should not be changed manually.

Manual deployment (⚠️ handle with care)

This project uses two GitHub dispatch workflows to manage manual deployments: dispatch-apps.yml and dispatch-infrastructure.yml. These workflows allow for manual triggers of deployments through GitHub Actions, providing flexibility for deploying specific versions to designated environments.

Using dispatch-apps.yml

The dispatch-apps.yml workflow is responsible for deploying applications. To trigger this workflow:

1. Navigate to the Actions tab in the GitHub repository.
2. Select the Dispatch Apps workflow.
3. Click on "Run workflow".
4. Fill in the required inputs:
   • environment: Choose the target environment (test, staging, or prod).
   • version: Specify the version to deploy. Could be git tag or a docker-tag published in packages.
   • runMigration (optional): Indicate whether to run database migrations (true or false).

This workflow will handle the deployment of applications based on the specified parameters, ensuring that the correct version is deployed to the chosen environment.

Using dispatch-infrastructure.yml

The dispatch-infrastructure.yml workflow is used for deploying infrastructure components. To use this workflow:

1. Go to the Actions tab in the GitHub repository.
2. Select the Dispatch Infrastructure workflow.
3. Click on "Run workflow".
4. Provide the necessary inputs:
   • environment: Select the environment you wish to deploy to (test, staging, or prod).
   • version: Enter the version to deploy, which should correspond to a git tag.

This workflow facilitates the deployment of infrastructure to the specified environment, using the version details provided.

GitHub Actions

Naming conventions for GitHub Actions:

• workflow-*.yml: Reusable workflows
• ci-cd-*.yml: Workflows that are triggered by an event
• dispatch-*.yml: Workflows that are dispatchable

The workflow-check-for-changes.yml workflow uses the tj-actions/changed-files action to check which files have been altered since last commit or tag. We use this filter to ensure we only deploy backend code or infrastructure if the respective files have been altered.

Infrastructure

Infrastructure definitions for the project are located in the .azure/infrastructure folder. To add new infrastructure components, follow the existing pattern found within this directory. This involves creating new Bicep files or modifying existing ones to define the necessary infrastructure resources.

For example, to add a new storage account, you would:

• Create or update a Bicep file within the .azure/infrastructure folder to include the storage account resource definition.
• Ensure that the Bicep file is referenced correctly in .azure/infrastructure/infrastructure.bicep to be included in the deployment process.

Refer to the existing infrastructure definitions as templates for creating new components.

Deploying a new infrastructure environment

A few resources need to be created before we can apply the Bicep to create the main resources.

The resources refer to a source key vault in order to fetch the necessary secrets and store them in the key vault for the environment. An ssh-key is also necessary for the ssh-jumper used to access the resources in Azure within the vnet.

Use the following steps:

• Ensure a source key vault exist for the new environment. Either create a new key vault or use an existing key vault. Currently, two key vaults exist for our environments. One in the test subscription used by Test and Staging, and one in our Production subscription, which Production uses. Ensure you add the necessary secrets that should be used by the new environment. Read here to learn about secret convention Configuration Guide. Ensure also that the key vault has the following enabled: Azure Resource Manager for template deployment.

• Ensure that a role assignment Key Vault Secrets User and Contributer(should be inherited) is added for the service principal used by the GitHub Entra Application.

• Create an SSH key in Azure and discard the private key. We will use the az cli to access the virtual machine so storing the ssh key is only a security risk.

• Create a new environment in GitHub and add the following secrets: AZURE_CLIENT_ID, AZURE_SOURCE_KEY_VAULT_NAME, AZURE_SOURCE_KEY_VAULT_RESOURCE_GROUP, AZURE_SOURCE_KEY_VAULT_SUBSCRIPTION_ID, AZURE_SUBSCRIPTION_ID, AZURE_TENANT_ID and AZURE_SOURCE_KEY_VAULT_SSH_JUMPER_SSH_PUBLIC_KEY

• Add a new file for the environment .azure/infrastructure/<env>.bicepparam. <env> must match the environment created in GitHub.

• Add the new environment in the dispatch-infrastructure.yml list of environments.

• Run the GitHub action Dispatch infrastructure with the version you want to deploy and environment. All the resources in .azure/infrastructure/main.bicep should now be created.

• (The GitHub action might need to restart because of a timeout when creating Redis).

Connecting to resources in Azure

There is a ssh-jumper virtual machine deployed with the infrastructure. This can be used to create a ssh-tunnel into the vnet. There are two ways to establish connections:

1. Using az ssh commands directly:

   # Connect to the VNet using:
   az ssh vm --resource-group dp-be-<env>-rg --vm-name dp-be-<env>-ssh-jumper
   
   # Or create an SSH tunnel for specific resources (e.g., PostgreSQL database):
   az ssh vm -g dp-be-<env>-rg -n dp-be-<env>-ssh-jumper -- -L 5432:<database-host-name>:5432

   This example forwards the PostgreSQL default port (5432) to your localhost. Adjust the ports and hostnames as needed for other resources.

   You may be prompted to install the ssh extension.

2. Using the forwarding utility script:

   See scripts/forward-bash/README.md for a more user-friendly way to establish database connections through SSH.

Applications

All application Bicep definitions are located in the .azure/applications folder. To add a new application, follow the existing pattern found within this directory. This involves creating a new folder for your application under .azure/applications and adding the necessary Bicep files (main.bicep and environment-specific parameter files, e.g., test.bicepparam, staging.bicepparam).

For example, to add a new application named web-api-new, you would:

• Create a new folder: .azure/applications/web-api-new
• Add a main.bicep file within this folder to define the application's infrastructure.
• Use the appropriate Bicep-modules within this file. There is one for Container apps which you most likely would use.
• Add parameter files for each environment (e.g., test.bicepparam, staging.bicepparam) to specify environment-specific values.

Refer to the existing applications like web-api-so and web-api-eu as templates.

Deploying applications in a new infrastructure environment

Ensure you have followed the steps in Deploying a new infrastructure environment to have the resources required for the applications.

Use the following steps:

• From the infrastructure resources created, add the following GitHub secrets in the new environment (this will not be necessary in the future as secrets would be added directly from infrastructure deployment): AZURE_APP_CONFIGURATION_NAME, AZURE_APP_INSIGHTS_CONNECTION_STRING, AZURE_CONTAINER_APP_ENVIRONMENT_NAME, AZURE_ENVIRONMENT_KEY_VAULT_NAME, AZURE_REDIS_NAME, AZURE_RESOURCE_GROUP_NAME, AZURE_SERVICE_BUS_NAMESPACE_NAME and AZURE_SLACK_NOTIFIER_FUNCTION_APP_NAME

• Add new parameter files for the environment in all applications .azure/applications/*/<env>.bicepparam

• Run the GitHub action Dispatch applications in order to deploy all applications to the new environment.

• To expose the applications through APIM, see Common APIM Guide