Error calling ne: incompatible types for comparison

[Archer-Li]

template: k8s-monitoring/templates/validations.yaml:5:4: executing "k8s-monitoring/templates/validations.yaml" at <include "destinations.validate" .>: error calling include: template: k8s-monitoring/templates/destinations/_destination_validations.tpl:40:46: executing "destinations.validate" at <ne $destination.protocol "http">: error calling ne: incompatible types for comparison

I think (ne $destination.protocol "http") will be evaluated for every destination item (including loki, prometheus...etc.). Without protocol set will immediately causing mismatch type error.

{{- if and (eq $destination.type "otlp") (ne $destination.protocol "http") (regexMatch "otlp-gateway-.+grafana\\.net" $destination.url) }}

Helm version: 2.0.10
Helm value.yaml:

destinations:
  - name: my-prom
    type: prometheus
    auth:
      type: basic
  - name: my-logs
    type: loki
    auth:
      type: basic
clusterMetrics:
  enabled: true
clusterEvents:
  enabled: true
podLogs:
  enabled: true
integrations:
  alloy:
    instances:
      - name: alloy
        labelSelectors:
          app.kubernetes.io/name:
            - alloy-metrics
            - alloy-singleton
            - alloy-logs
alloy-metrics:
  enabled: true
alloy-singleton:
  enabled: true
alloy-logs:
  enabled: true

[Archer-Li]

Also is there a way to disable all validation? Thanks for answering!

[petewall]

Please try updating your Helm cli. At minimum, we use helm version 3.14.

[jewbetcha]

@Archer-Li just checking in if you were able to update helm CLI, and if that helped

[Archer-Li]

Hello, We are using Terraform to installing K8s-helm charts. Here's the list of versions we were using:

required_providers {
    aws        = "~> 3.72.0"
    kubernetes = "~> 2.7.1"
    helm       = "~> 2.4.1"
    kubectl = {
      source  = "gavinbunney/kubectl"
      version = ">= 1.13.1"
    }
  }

We have attempted to update helm version with latest: 3.0.0-pre1, but not getting any luck yet

[petewall]

Let me check with my own terraform deployed instance. Can you share the helm_resource definition?

[Archer-Li]

Sure, here's the definition:

resource "helm_release" "grafana-k8s-monitoring" {
  name             = "grafana-k8s-monitoring"
  repository       = "https://grafana.github.io/helm-charts"
  chart            = "k8s-monitoring"
  namespace        = var.namespace
  create_namespace = true
  atomic           = true
  timeout          = 900
  force_update     = true
  recreate_pods    = false
  cleanup_on_fail  = true

  values = [file("${path.module}/values.yaml")]

  set {
    name  = "cluster.name"
    value = var.cluster_name
  }
  
  More set blocks for secrets only...
}

We do have the helm with same definition deployed successfully on our linux/arm64 cluster. But linux/amd64 is keeps giving us type check issues.

[petewall]

OK! I don't know why it works on some clusters and not others. I'll re-organize the checks to make sure that it doesn't check for protocol on non-otlp destinations.