404 when running terraform destroy for a kubernetes cluster and load balancer.

[surgiie]

I create the following cluster:

resource "digitalocean_kubernetes_cluster" "cluster" {
  name   = var.cluster_name
  region = var.region
  version = "1.28.2-do.0"
  auto_upgrade = true
  registry_integration = true

  node_pool {
    name       = "${var.cluster_name}-default-worker-pool"
    size       = var.node_size
    node_count = var.node_count
    auto_scale = false
  }

  depends_on = [
    digitalocean_loadbalancer.ingress_load_balancer,
  ]
}

And create the following load balancer for ingress

resource "digitalocean_loadbalancer" "ingress_load_balancer" {
  name   = "${var.cluster_name}-lb-1"
  region = var.region
  size = "lb-small"
  algorithm = "round_robin"

  forwarding_rule {
    entry_port     = 80
    entry_protocol = "http"

    target_port     = 80
    target_protocol = "http"
  }
  lifecycle {
    ignore_changes = [
        forwarding_rule,
    ]
  }
}

I also create the ingress nginx controller and associate the load balancer with it:

resource "helm_release" "nginx_ingress_chart" {
  name       = "nginx-ingress-controller"
  namespace  = "ingress-nginx"
  repository = "https://kubernetes.github.io/ingress-nginx"
  chart      = "ingress-nginx"
  set {
    name  = "service.type"
    value = "LoadBalancer"
  }
  set {
    name = "defaultBackend.enabled"
    value = false
  }
  set {
    name  = "controller.service.annotations.kubernetes\\.digitalocean\\.com/load-balancer-id"
    value = digitalocean_loadbalancer.ingress_load_balancer.id
  }
  depends_on = [
    digitalocean_loadbalancer.ingress_load_balancer,
  ]
}

Resources are created without issues as expected, but when I do a terraform destroy, it destroys some resources just fine but fails during the load balancer delete as it fails with the following error:

│ Error: Error deleting Load Balancer: DELETE https://api.digitalocean.com/v2/load_balancers/5eeb187b-3242-412b-b118-0028245d9946: 404 (request "4e0a4054-299b-40c4-b37e-d6f89a053d85") Load Balancer not found

It feels like digital ocean cascades the deletion of the cluster to the load balancer so when it tries to delete the load balancer, it is already gone. Any ideas on how to fix this or am I doing something wrong on my end?

[andrewsomething]

Thanks for flagging this issue. I can reproduce this as well. We should be able to safely ignore a 404 error on load balancer delete. This PR should resolve the issue: #1067 It will be included in the next release.

[surgiie]

Perfect thanks for your work on this! 👍🏽