错误 kube-system/configmaps：拨打 tcp 127.0.0.1:80：连接：连接被拒绝

Question

我正在尝试部署具有自我管理节点组的集群。无论我使用什么配置选项，我总是会出现以下错误：

\n

\n

错误：发布\n“ http://localhost/api/v1/namespaces/kube-system/configmaps ”：\ndial tcp 127.0.0.1:80：连接：连接被拒绝\nmodule.eks-ssp.kubernetes_config_map.aws_auth[ 0]on\n.terraform/modules/eks-ssp/aws-auth-configmap.tf 第 19 行，在资源\n"kubernetes_config_map" "aws_auth":resource\n"kubernetes_config_map" "aws_auth" {

\n

\n

\xe2\x80\x8b

\n

该.tf文件如下所示：

\n

module "eks-ssp" {\nsource = "github.com/aws-samples/aws-eks-accelerator-for-terraform"\n\n# EKS CLUSTER\ntenant            = "DevOpsLabs2"\nenvironment       = "dev-test"\nzone              = ""\nterraform_version = "Terraform v1.1.4"\n\n# EKS Cluster VPC and Subnet mandatory config\nvpc_id             = "xxx"\nprivate_subnet_ids = ["xxx","xxx", "xxx", "xxx"]\n\n# EKS CONTROL PLANE VARIABLES\ncreate_eks         = true\nkubernetes_version = "1.19"\n\n# EKS SELF MANAGED NODE GROUPS\nself_managed_node_groups = {\nself_mg = {\nnode_group_name        = "DevOpsLabs2"\nsubnet_ids             = ["xxx","xxx", "xxx", "xxx"]\ncreate_launch_template = true\nlaunch_template_os     = "bottlerocket"       # amazonlinux2eks  or bottlerocket or windows\ncustom_ami_id          = "xxx"\npublic_ip              = true                   # Enable only for public subnets\npre_userdata           = <<-EOT\nyum install -y amazon-ssm-agent \\\nsystemctl enable amazon-ssm-agent && systemctl start amazon-ssm-agent \\\nEOT\n\ndisk_size     = 20\ninstance_type = "t2.small"\ndesired_size  = 2\nmax_size      = 10\nmin_size      = 2\ncapacity_type = "" # Optional Use this only for SPOT capacity as  capacity_type = "spot"\n\nk8s_labels = {\nEnvironment = "dev-test"\nZone        = ""\nWorkerType  = "SELF_MANAGED_ON_DEMAND"\n}\n\nadditional_tags = {\nExtraTag    = "t2x-on-demand"\nName        = "t2x-on-demand"\nsubnet_type = "public"\n}\ncreate_worker_security_group = false # Creates a dedicated sec group for this Node Group\n},\n}\n}\n\nmodule "eks-ssp-kubernetes-addons" {\nsource = "github.com/aws-samples/aws-eks-accelerator-for-terraform//modules/kubernetes-addons"\n\neks_cluster_id                        = module.eks-ssp.eks_cluster_id\n\n# EKS Addons\nenable_amazon_eks_vpc_cni             = true\nenable_amazon_eks_coredns             = true\nenable_amazon_eks_kube_proxy          = true\nenable_amazon_eks_aws_ebs_csi_driver  = true\n\n#K8s Add-ons\nenable_aws_load_balancer_controller   = true\nenable_metrics_server                 = true\nenable_cluster_autoscaler             = true\nenable_aws_for_fluentbit              = true\nenable_argocd                         = true\nenable_ingress_nginx                  = true\n\ndepends_on = [module.eks-ssp.self_managed_node_groups]\n}\n

\n

提供商：

\n

terraform {\n\n  backend "remote" {}\n\n  required_providers {\n    aws = {\n      source  = "hashicorp/aws"\n      version = ">= 3.66.0"\n    }\n    kubernetes = {\n      source  = "hashicorp/kubernetes"\n      version = ">= 2.6.1"\n    }\n    helm = {\n      source  = "hashicorp/helm"\n      version = ">= 2.4.1"\n    }\n  }\n}\n

\n

Answer 1

根据 Github 存储库 [1] 中提供的示例，我的猜测是provider缺少配置块，无法按预期工作。查看问题中提供的代码，似乎需要添加以下内容：

data "aws_region" "current" {}

data "aws_eks_cluster" "cluster" {
  name = module.eks-ssp.eks_cluster_id
}

data "aws_eks_cluster_auth" "cluster" {
  name = module.eks-ssp.eks_cluster_id
}

provider "aws" {
  region = data.aws_region.current.id
  alias  = "default" # this should match the named profile you used if at all
}

provider "kubernetes" {
  experiments {
    manifest_resource = true
  }
  host                   = data.aws_eks_cluster.cluster.endpoint
  cluster_ca_certificate = base64decode(data.aws_eks_cluster.cluster.certificate_authority.0.data)
  token                  = data.aws_eks_cluster_auth.cluster.token
}

如果helm还需要，我认为还需要添加以下块[2]：

provider "helm" {
  kubernetes {
    host                   = data.aws_eks_cluster.cluster.endpoint
    token                  = data.aws_eks_cluster_auth.cluster.token
    cluster_ca_certificate = base64decode(data.aws_eks_cluster.cluster.certificate_authority.0.data)
  }
}

kubernetes和的提供程序参数参考helm分别在 [3] 和 [4] 中。

[1] https://github.com/aws-samples/aws-eks-accelerator-for-terraform/blob/main/examples/eks-cluster-with-self-management-node-groups/main.tf#L23 -L47

[2] https://github.com/aws-samples/aws-eks-accelerator-for-terraform/blob/main/examples/eks-cluster-with-eks-addons/main.tf#L49-L55

[3] https://registry.terraform.io/providers/hashicorp/kubernetes/latest/docs#argument-reference

[4] https://registry.terraform.io/providers/hashicorp/helm/latest/docs#argument-reference