app-helm-chart [Archived]

This repository is archived and no longer maintained. All helm charts have been moved to the DND-IT/helm-charts repository.

You can find how to update to the new repository here

Contains the generic helm chart for an application. It is used by Prometheus & Disco project

Usage

Reference the release of the chart you want to deploy in terraform

resource "helm_release" "app" {
  chart     = "https://github.com/DND-IT/app-helm-chart/archive/3.3.2.tar.gz"
  values = [
    templatefile("values.yaml")
  ]
  set {
    name  = "service.name"
    value = "myname-myenv"
  }
  set {
    name  = "scale.enabled"
    value = "false"
  }
}

AWS Role

If you specify a IAM role in aws_iam_role_arn a service account will be created to take advantage of fine grained permission for pods

Pod Disruption Budget

By default this chart include a sort of fail-safe mechanism to prevent huge application disruptions. K8s documentation explains very well what are the benefits and the limits of settings PDBs in case of voluntary disruptions.

For highly-available deployments, the chart allows a minimum availability of 50%. For production-critical applications we recommend evaluating whether 50% is enough.

You can configure it by overriding the scale.minAvailable parameter. (e.g. 90%)

Note: PDBs are only deployed together with deployments that have more than one replica.

Tests

On PR, a github workflow is run which tests the chart with *.yaml value files in the ci directory. And it will try to deploys the app-helm-chart on a k8s cluster using various *-value.yaml value files in the ci directory). Workflow is copied from https://github.com/renovatebot/helm-charts.

Releases

No automatic release workflow. Once a change is merged on master and new tag / release have to be created in the github console