kargo-workshop

Steps for the workshop

1. Go to this repository (Click here!) and fork it.
   Copy the URL (it should end with .git) of this forked repo, as you'll need it in step 4.

2. Start the dev container, which can be done in 2 ways:

   • Open GitHub, clone this repo with one of the following commands:

     # https
     git clone https://github.com/fullstaq-labs/edgecase2024-kargo-workshop.git
     
     # ssh
     git clone git@github.com:fullstaq-labs/edgecase2024-kargo-workshop.git

     Open the folder where you've cloned the repo locally in Visual Studio Code and follow the instructions from the image below.
     

Note

Don't forget to stop and delete your Codespace afterwards!

• Go to this Repo in GitHub and open Codespaces (follow the instructions from the image below).
  

3. When the container starts, the following tools are installed:

   • Kind cluster
     Kargo needs to run on a Kubernetes cluster. Therefore, we will create a Kubernetes cluster using kind.
   • Cert-Manager
     Cert-manager is used for self-signing a certificate used to identify Kargo's webhook server to the Kubernetes API server.
   • ArgoCD
     To deploy applications to the Kubernetes cluster, ArgoCD needs to be available.
   • Argo-Rollouts
     To use AnalysisTemplate resources, Argo Rollouts is needed.
   • Kargo
     Obviously, Kargo is needed for this workshop :-)

Note

• Task is used to orchestrate all these scripts.
• Gum is used for structured logs and a user-like interface.

4. Run the following command in the container (the value should be the url of the repo you just forked.):

   export GITOPS_REPO_URL="<Replace this value>.git"

5. If you don't have a GitHub token, start creating one. This needs to be a classic token; creation is explained in this guide.

Basic Scenario

With these next steps, we will start the basic scenario.

6. We are now going to apply the configuration in the Kubernetes cluster.

   To make it easy for you, I've prepared some manifests, which can be found in the folder demo-basic.

   The following parts are created:

   • ArgoCD ApplicationSet
   • Kargo Project
   • Kubernetes secret containing the repository credentials
   • Kargo Warehouse
   • Kargo Stages

   To apply the changes, I've created a taskfile that will inject certain values. You can apply the code with the following command:

   task init_basic

Note

Have a look at the manifests that are applied, which are in the folder demo-basic. Important for the flow are the stages. To get a better understanding of the flow, take a look at the promotionMechanisms. Look for the difference between the test and other stages.

7. We are now going to have our first experience with the Kargo CLI (which is preinstalled in this environment).

   First, let's check if the Kargo CLI is working correctly:

   kargo

   The result should look like this:

   Usage:
   kargo [flags]
   kargo [command]
   
   Available Commands:
   apply       Apply a resource from a file or from stdin
   approve     Manually approve a piece of freight for promotion to a stage
   completion  Generate the autocompletion script for the specified shell
   config      Manage Kargo CLI configuration
   create      Create a resource from a file or from stdin
   dashboard   Open the Kargo Dashboard in your default browser
   delete      Delete resources by file and names
   get         Display one or many resources
   grant       Grant a role to a user or grant permissions to a role
   help        Help about any command
   login       Log in to a Kargo API server
   logout      Log out of the Kargo API server
   promote     Promote a piece of freight
   refresh     Refresh a stage or warehouse
   revoke      Revoke a role from a user or revoke permissions from a role
   server      Start a local Kargo API server
   update      Update a resource
   verify      Verify a stage
   version     Show the client and server version information
   
   Flags:
   -h, --help   help for kargo
   
   Use "kargo [command] --help" for more information about a command.

   To interact with the environment, we need to log in to Kargo:

   kargo login https://localhost:31444 --admin --password=admin --insecure-skip-tls-verify

   After that, let's set the Project for the next steps using the following command:

   kargo config set-project "demo-basic"

   Now let's start reviewing our applied configuration with some commands.

   Check the Projects:

   kargo get projects

   Check the Warehouses:

   kargo get warehouses

   Check the Stages:

   kargo get stages

   Check the Freight:

   kargo get freight

   All these commands should show you some output.

8. Now we have a solid base to work with.

   Let's start promoting the test Stage. This promotion makes it possible to promote the image from the piece of Freight to the Stage.

   To do this, we first need to know which freight is available with the following command:

   kargo get freight

   Copy the "alias", which we need in the next command:

   kargo promote --stage=test --freight-alias=<Freight alias>
   
   ## Use the following command to follow the status (exit with Ctrl + C)
   watch kargo get promotions

9. It's not only possible to view the result on the command line, but also in the Kargo Dashboard.

   Normally you can open this with the command:

   kargo dashboard

   Unfortunately, this doesn't work in the dev container, so please click here (ignore certificate errors, just proceed to the website).

   Login with user admin and password admin and click on the project demo-basic.

10. After the promotion, a process in ArgoCD is started to deploy the application through ArgoCD. To view this process, go to ArgoCD by clicking here (ignore certificate errors, just proceed to the website).

    Login with user admin and password admin.

    If you want to see progress from the application, click on the test Application. You should be able to see a fully deployed application.

11. Now that I've explained how you can handle the promotion to Stages, you should be able to do it yourself.

    Before you start promoting in the normal way, first try to directly promote to prod. You should receive an error, but there is a way of working around this (in case of emergency).

    Kargo has the possibility to approve a piece of Freight for a certain Stage. This can be done using the following command:

    kargo approve --stage prod --freight-alias <Freight alias>

12. Once done, your Kargo dashboard should look something like:

    

    ArgoCD should look like:

    

13. To verify, you can view all workloads running on the workshop cluster. Just run the following command:

    kubectl get pods -A -l "project=demo-basic"

Advanced Scenario

14. We are now going to apply the configuration of the advanced scenario in the Kubernetes cluster. To make it easy for you, I've prepared some manifests, which can be found in the folder demo-advanced.

    The following parts are created:

    • ArgoCD ApplicationSet
    • Kargo Project
    • Kubernetes secret containing the repository credentials
    • Kargo Warehouse
    • Kargo Stages

    To apply the changes, I've created a taskfile that will inject certain values. You can apply the code with the following command:

    task init_advanced

    As you can see, this scenario has many more Stages and different dependencies. A new element, AnalysisTemplate resources, is also introduced, which we will use later in this scenario.

Important

Have a look at the manifests that are applied, which are in the folder demo-advanced. Compared to the basic demo, there are more changes than just the number of stages. The kargo.yaml file now also has analysistemplates, which you will see later during the workshop.

15. Let's check if everything is deployed correctly!

    First, set the correct Project:

    kargo config set-project "demo-advanced"

    Now let's start reviewing our applied configuration with some commands.

    Check the Projects (you will now see 2 of them):

    kargo get projects

    Check the Warehouses:

    kargo get warehouses

    Check the Stages:

    kargo get stages

    Check the Freight:

    kargo get freight

    All these commands should show you some output.

16. As we did with the basic scenario, we should promote the test Stage.

    In order to do that, we first need to get the Freight (which differs from the basic demo):

    kargo get freight

    After that, copy the "alias" which we need in the next command:

    kargo promote --stage=test --freight-alias=<Freight alias>
    
    ## Use the following command to follow the status (exit with Ctrl + C)
    watch kargo get promotions

Note

Feel free to have a look at the dashboard. Now you should see two projects; choose demo-advanced.

17. As in the basic demo, we also have to promote the other Stages.

    Let's start with promoting to uat:

    kargo promote --stage=uat --freight-alias=<Freight alias>
    
    ## Use the following command to follow the status (exit with Ctrl + C)
    watch kargo get promotions

18. Now let's promote to pre-prod-b (not pre-prod-a!):

    kargo promote --stage=pre-prod-b --freight-alias=<Freight alias>
    
    ## Use the following command to follow the status (exit with Ctrl + C)
    watch kargo get promotions

    There is a big difference here compared to the basic demo.

    An AnalysisTemplate runs in the background and should prevent promotion from pre-prod-b to prod-canary.

    Let's try this:

    kargo promote --stage=prod-canary --freight-alias=<Freight alias>
    
    ## Use the following command to follow the status (exit with Ctrl + C)
    watch kargo get promotions

    If you haven't done this, please have a look at the AnalysisTemplate. This can be done through the manifests in this repository and also through the Kargo UI.

19. After this, promote to pre-prod-a:

    kargo promote --stage=pre-prod-a --freight-alias=<Freight alias>
    
    ## Use the following command to follow the status (exit with Ctrl + C)
    watch kargo get promotions

    Now you should be able to promote to the other Stages. So promote to prod-canary and prod-ga.

20. To verify, you can view all workloads running on the workshop cluster.

    Just run the following command:

    kubectl get pods -A -l "project=demo-advanced"

21. Once done, your Kargo dashboard should look something like:

    

    ArgoCD should look like:

    

22. Now we've completed the entire flow 🎉!

    We are now going to create a new piece of Freight. This can be done by changing the version of the value of semverConstraint for the Warehouse (in this file):

    apiVersion: kargo.akuity.io/v1alpha1
    kind: Warehouse
    metadata:
        name: ${KARGO_PROJECT}
        namespace: ${KARGO_PROJECT}
    spec:
        subscriptions:
        - image:
            repoURL: nginx
            semverConstraint: <new SemVer constraint, i.e. "*" or ">=1.27.0">

23. After completing the change, let's apply the change:

    task init_advanced

24. When done, start promoting all Stages:

    • test
    • uat
    • pre-prod-a
    • pre-prod-b
    • prod-canary
    • prod-ga