ArgoKit

ArgoKit is a set of reusable jsonnet templates that makes it easier to deploy ArgoCD applications on SKIP. It is a work in progress and will be updated as we learn more about how to best deploy with ArgoCD on SKIP. If you have any questions, please reach out to the #gen-argocd channel in Slack.

Installation

Assuming you have followed the Getting Started guide, you can use ArgoKit in your apps-repo. The first step is to include the ArgoKit library by running the following command:

$ git submodule add https://github.com/kartverket/argokit.git

Alternatively you can use jsonnet-bundler if prefer package managers over submodules. To do this, install the CLI by following the instructions in the jsonnet-bundler repo and run the following command:

$ jb install https://github.com/kartverket/argokit@main

Automatic version updates

It is highly recommended to use dependabot to automatically update the ArgoKit version when a new version is released. To do this, add the following to your .github/dependabot.yml file:

version: 2
updates:
  - package-ecosystem: git-submodules
    directory: /
    schedule:
      interval: daily

With this configuration, dependabot will automatically create a PR to update the ArgoKit version once a day provided that a new version has been released.

Version tracking

ArgoKit automatically tracks version information in deployed resources. When developing locally or testing branches, run ./generate-version.sh to update version metadata. See VERSION_GENERATION.md for details.

Usage with jsonnet

If you use jsonnet in your apps-repo, you can use the ArgoKit library to deploy ArgoCD applications by including the argokit.libsonnet file in your jsonnet file and calling the argokit.Application function. For example, to deploy an application, you can use the following jsonnet file:

local argokit = import 'argokit/jsonnet/argokit.libsonnet';

local Probe = {
    path: "/healthz",
    port: 8080,
    failureThreshold: 3,
    timeout: 1,
    initialDelay: 0,
};

local BaseApp = {
    spec: {
        port: 8080,
        replicas: {
            min: 2,
            max: 5,
            targetCPUUtilization: 80,
        },
        liveness: Probe,
        readiness: Probe,
    },
};

[
    BaseApp + argokit.Application("foo-backend") {
        spec+: {
            image: "hello-world",
            ingresses: ["foo.bar.com"],
            accessPolicy: {
                inbound: {
                    rules: [{
                        application: "foo-frontend",
                    }],
                },
            },
        },
    },
]

jsonnet argokit API

The following templates are available for use in the argokit.libsonnet file:

Template	Description	Example
argokit.Application	Creates a Skiperator application	See above
argokit.GSMSecretStore	Creates a Google Secret Manager External Secrets SecretStore	examples/jsonnet/secretstore-gsm.jsonnet
argokit.GSMSecret	Creates a Google Secret Manager External Secrets Secret	examples/jsonnet/secretstore-gsm.jsonnet
argokit.Roles	Creates a set of RBAC roles for this namespace	examples/jsonnet/roles.jsonnet
argokit.AccessPolicies	Configures inbound and outbound access policies for services and jobs	examples/jsonnet/accessPolicies.jsonnet

argoKit's Environment API

Template	Description	Example
argokit.Env.variable	Creates environment variables for an app	examples/jsonnet/envVariables.jsonnet
argokit.Env.variableSecret	Creates environment variable from a secret	examples/jsonnet/envVariables.jsonnet
argokit.Env.variableSecretJob	Creates environment variable from a secret for a Job container	examples/jsonnet/envVariables.jsonnet

---

The following templates are available for use in the dbArchive.libsonnet file:

Template	Description	Example
dbArchive.dbArchiveJob	Creates a SKIPJob that creates a sql dump and stores it in S3	examples/jsonnet/dbArchive.jsonnet

Input parameters

dbArchiveJob

Parameter	Type	Default Value	Description
instanceName	String	-	Required. A unique name for the job and related resources. This name is used as a base for SKIPJob and secrets.
schedule	String	-	Required. A cron string defining when the job should run (e.g., "0 2 * * *" to run at 02:00 every night).
databaseIP	String	-	Required. The IP address of the PostgreSQL database to be archived.
gcpS3CredentialsSecret	String	-	Required. Name of the secret in GSM containing S3 credentials (AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY).
databaseName	String	-	Required. Name of the database to be archived.
archiveUser	String	'postgres'	The database user the job should use to connect.
serviceAccount	String	'dummyaccount@gcp.iam'	GCP Service Account used by the Kubernetes job to authenticate against Google Cloud (e.g., to fetch secrets from GSM).
cloudsqlInstanceConnectionName	String	-	Required. The connection name of the Cloud SQL instance (format: project:region:instance). Needed for Cloud SQL Auth Proxy.
port	Integer	5432	The port number of the PostgreSQL database.
S3Host	String	's3-rin.statkart.no'	The hostname of the S3 endpoint where the archive should be stored.
S3DestinationPath	String	-	Required. Full S3 path where the database archive should be placed (e.g., s3://my-bucket/archive/database/).
fullDump	Bool	false	Flag to include database roles without passwords in the dump.

Contributing

Contributions are welcome! Please open an issue or PR if you would like to see something changed or added.

License

ArgoKit is licensed under the MIT License. See LICENSE for the full license text.