helm-chart/README.md
xeruf 8300c7599f
All checks were successful
continuous-integration/drone/push Build is passing
docs: re-add removed documentation (#19)
Co-authored-by: xeruf <27jf@pm.me>
Reviewed-on: #19
Co-authored-by: xeruf <git@jfischer.org>
Co-committed-by: xeruf <git@jfischer.org>
2024-02-09 18:14:49 +00:00

6.3 KiB

Vikunja Helm Chart

This Helm Chart deploys both the Vikunja frontend and Vikunja api containers, in addition to other Kubernetes resources so that you'll have a fully functioning Vikunja deployment quickly. Also, you can deploy Bitnami's PostgreSQL and Redis as subcharts if you want, as Vikunja can utilize them as its database and caching mechanism (respectively).

See https://artifacthub.io/packages/helm/vikunja/vikunja for version information and installation instructions.

Quickstart

Define ingress settings according to your controller (for both API and Frontend) to access the application. You can set all Vikunja API options as yaml under api.configMaps.config.data.config.yml: https://vikunja.io/docs/config-options

For example, you can disable registration (if you do not with to allow others to register on your Vikunja), by providing the following values in your values.yaml:

api:
  configMaps:
    config:
      enabled: true
      data:
        config.yml:
          service:
            enableregistration: false

You can still create new users by executing the following command in the api container:

./vikunja user create --email <user@email.com> --user <user1> --password <password123>

Advanced Features

Replicas

To effectively run multiple replicas of the API, make sure to set up the redis cache as well by setting api.configMaps.config.data.config.yml.keyvalue.type to redis, configuring the redis subchart (see values.yaml) and the connection in Vikunja

Use an existing file volume claim

In the values.yaml file, you can either define your own existing Persistent Volume Claim (PVC) or have the chart create one on your behalf.

To have the chart use your pre-existing PVC:

api:
  persistence:
    data:
      enabled: true
      existingClaim: <your-claim>

To have the chart create one on your behalf:

# You can find the default values 
api:
  enabled: true
  persistence:
    data:
      enabled: true
      accessMode: ReadWriteOnce
      size: 10Gi
      mountPath: /app/vikunja/files
      storageClass: storage-class

Utilizing environment variables from Kubernetes secrets

Each environment variable that is "injected" into a pod can be sourced from a Kubernetes secret. This is useful when you wish to add values that you would rather keep as secrets in your GitOps repo as environment variables in the pods.

Assuming that you had a Kubernetes secret named vikunja-env, this is how you would add the value stored at key VIKUNJA_DATABASE_PASSWORD as the environment variable named VIKUNJA_DATABASE_PASSWORD:

api:
  env:
    VIKUNJA_DATABASE_PASSWORD:
      valueFrom:
        secretKeyRef:
          name: vikunja-env
          key: VIKUNJA_DATABASE_PASSWORD
    VIKUNJA_DATABASE_USERNAME: "db-user"

If the keys within the secret are the names of environment variables, you can simplify passing multiple values to this:

api:
  envFrom:
    - secretRef:
      name: vikunja-secret-env
  env:
    VIKUNJA_DATABASE_USERNAME: "db-user"

This will add all keys within the Kubernetes secret named vikunja-secret-env as environment variables to the api pod. Additionally, if you did not have the key VIKUNJA_DATABASE_USERNAME in the vikunja-secret-env secret, you could still define it as an environment variable seen above.

How the envFrom key works can be seen here.

Utilizing a Kubernetes secret as the config.yml file instead of a ConfigMap

If you did not wish to use the ConfigMap provided by the chart, and instead wished to mount your own Kubernetes secret as the config.yml file in the api pod, you could provide values such as the following (assuming asdf-my-custom-secret1 was the name of the secret that had the config.yml file):

api:
  persistence:
    config:
      type: secret
      name: asdf-my-custom-secret1

Then your secret should look something like the following so that it will mount properly:

apiVersion: v1
kind: Secret
metadata:
  name: asdf-my-custom-secret1
  namespace: vikunja
type: Opaque
stringData:
  config.yml: |
    key1: value1
    key2: value2
    key3: value3    

Modifying Deployed Resources

Oftentimes, modifications need to be made to a Helm chart to allow it to operate in your Kubernetes cluster. Anything you see in bjw-s' common library, including the top-level keys, can be added and subtracted from this chart's values.yaml, underneath the api, frontend, and (optionally) typesense key.

For example, if you wished to create a serviceAccount as can be seen here for the api pod:

api:
  serviceAccount: 
    create: true

Then, (for some reason), if you wished to deploy the frontend as a DaemonSet (as can be seen here), you could do the following:

frontend:
  controller:
    type: daemonset

Publishing

The following steps are automatically performed when a git tag for a new version is pushed to the repository. They are only listed here for reference.

  1. Pull all dependencies before packaging.
helm dependency update
  1. In order to publish the chart, you have to either use curl or helm cm-push.
helm package .
curl --user '<username>:<password>' -X POST --upload-file './<archive>.tgz' https://kolaente.dev/api/packages/vikunja/helm/api/charts
helm package .
helm repo add --username '<username>' --password '<password>' vikunja https://kolaente.dev/api/packages/vikunja/helm
helm cm-push './<archive>.tgz' vikunja

As you can see, you do not have to specify the name of the repository, just the name of the organization.