Taskfile.yml

# yaml-language-server: $schema=https://taskfile.dev/schema.json
version: '3'

vars:
  TF_DIR: "{{.ROOT_DIR}}/terraform"
  CONFIG_FILE: "{{.ROOT_DIR}}/config.yaml"
  STATE_DIR: "{{.ROOT_DIR}}/state"

includes:
  op:
    taskfile: ./op/Taskfile.yml
    dir: ./op
    vars:
      TF_DIR: "{{.TF_DIR}}"
      STATE_DIR: "{{.STATE_DIR}}"
  x:
    taskfile: ./extra/Taskfile.yml
    optional: true
    dir: ./extra
    vars:
      TF_DIR: "{{.TF_DIR}}"
      STATE_DIR: "{{.STATE_DIR}}"

tasks:
  default:
    cmds:
      - task --list
    silent: true
  download-images:
    desc: Download essential images
    cmds:
      - ./scripts/download-images.sh
      - ls -alh {{.ROOT_DIR}}/artifacts/images
  artifacts-up:
    desc: Build and run the artifacts server container
    cmds:
      - ./artifacts/run-server.sh
  artifacts-curl-test:
    desc: Curl artifacts server from the admin node
    cmds:
      - |
        ARTIFACTS_SERVER_URL="$(yq e '.artifact_server_url' {{.CONFIG_FILE}})"
        ssh -F {{.ROOT_DIR}}/state/ssh_config admin "sudo curl --version &>/dev/null || sudo apk add curl"
        ssh -F {{.ROOT_DIR}}/state/ssh_config admin "curl ${ARTIFACTS_SERVER_URL}"
  nfs-up:
    desc: Build and run the NFS server (nfs-ganesha) container
    cmds:
      - ./nfs/run-server.sh
  nfs-mount-test:
    desc: Mount the NFS export from the admin node
    preconditions:
      - sh: test -f {{.ROOT_DIR}}/state/ssh_config
        msg: "state/ssh_config not found, provision the admin node first"
      - sh: ssh -F {{.ROOT_DIR}}/state/ssh_config -o ConnectTimeout=5 admin true
        msg: "admin node is not reachable via SSH, make sure it is running"
    cmds:
      - |
        MGMT_IP="$(yq e '.networks.mgmt.bridge_ip' {{.CONFIG_FILE}})"
        ssh -F {{.ROOT_DIR}}/state/ssh_config admin "sudo mount.nfs4 -V &>/dev/null || sudo apk add nfs-utils"
        ssh -F {{.ROOT_DIR}}/state/ssh_config admin "sudo mkdir -p /mnt/nfs-test && sudo mount -t nfs4 ${MGMT_IP}:/backup_target /mnt/nfs-test"
        ssh -F {{.ROOT_DIR}}/state/ssh_config admin "sudo touch /mnt/nfs-test/.mount-test && sudo rm /mnt/nfs-test/.mount-test && sudo umount /mnt/nfs-test"
        echo "NFS mount test OK"
  generate-macs:
    desc: Generate MAC addresses for nodes and update config file
    cmds:
      - ./scripts/generate-mac-addresses.sh
  plan-networks:
    desc: Plan networks
    cmds:
      - ./scripts/plan-networks.sh {{.CLI_ARGS}}
  select-install-iso:
    desc: "Select an ISO and update config file. Use `-- VERSION` to specify a version directly, e.g. `task select-install-iso -- v1.3.0`"
    cmds:
      - ./artifacts/select-install-iso.sh {{.CLI_ARGS}}
  select-upgrade-iso:
    desc: "Select an ISO and update config file. Use `-- VERSION` to specify a version directly, e.g. `task select-upgrade-iso -- v1.3.0`"
    cmds:
      - ./artifacts/select-upgrade-iso.sh {{.CLI_ARGS}}
  init_tf:
    desc: Initialize Terraform and download provider plugins
    cmds:
      - terraform -chdir={{.TF_DIR}} init
    status:
      - test -f {{.TF_DIR}}/.terraform.lock.hcl
  init_keys:
    desc: Generate shared SSH key pairs if they don't already exist
    cmds:
      - mkdir -p state/keys
      - test -f state/keys/id_rsa || ssh-keygen -t rsa -b 4096 -f state/keys/id_rsa -N ""
    status:
      - test -f state/keys/id_rsa
  init:
    deps: [init_tf, init_keys]
  config-check:
    desc: Validate config.yaml for duplicate ip, mac, and vnc_port in nodes
    cmds:
      - python3 scripts/validate-config.py {{.CONFIG_FILE}}
  check:
    desc: Run all checks
    deps: [config-check]
  plan:
    desc: Plan the terraform deployment
    deps: [init]
    cmds:
      - terraform -chdir={{.TF_DIR}} plan
  build:
    desc: Build the Dev Infra
    deps: [init, check]
    cmds:
      - terraform -chdir={{.TF_DIR}} apply --auto-approve
  admin-up:
    desc: Power on admin node and wait for it to be ready
    cmds:
      - task: build
      - task: op:admin-poweron
      - task: op:admin-wait
      - task: op:admin-enable-egress-if-configured
  up:
    desc: Build base dev infra and install Harvester on nodes
    deps: [admin-up]
    cmds:
      - task: op:nodes-install
  rancher-install:
    desc: Bring up the Rancher VM and wait for it to be ready
    deps: [init]
    cmds:
      - task: op:rancher-install
  rancher-reinstall:
    desc: Reset the Rancher VM to initial state and provisiong a new Rancher Manager
    deps: [init]
    cmds:
      - task: op:rancher-reinstall
  nodes-install:
    desc: Power on all nodes and wait for Harvester cluster to be ready
    deps: [init]
    cmds:
      - task: op:nodes-install
  nodes-reinstall:
    desc: Reset node disks and re-install Harvester on all nodes again
    deps: [init]
    cmds:
      - task: op:nodes-reinstall
  clean:
    desc: Tear down the Dev Infra and clean up resources
    cmds:
      - echo "This will clean state files, destroy VMs, and tear down the terraform infra" && ./op/confirm-op.sh {{.CLI_ARGS}}
      - task: op:rancher-clean-state
      - task: op:nodes-destroy
      - task: op:nodes-clean-state
      - terraform -chdir={{.TF_DIR}} destroy --auto-approve
#      - rm -rf {{.STATE_DIR}}/keys
      - rm -f {{.STATE_DIR}}/ssh_config
      - rm -f kubeconfig
      - task: op:harvester-clean-state
  reset:
    cmds:
      - task clean
      - task build
  ssh:
    desc: SSH into VMs using generated SSH config. By default, it will ssh into admin node. Use `-- NODE` to specify a node, e.g. `task ssh -- node1`
    cmds:
      - ssh -F {{.ROOT_DIR}}/state/ssh_config {{if .CLI_ARGS}}{{.CLI_ARGS}}{{else}}admin{{end}}
  node-count:
    desc: Get or set the number of nodes in config.yaml. Use `-- N` to set the count, e.g. `task node-count -- 3`
    cmds:
      - ./op/nodes/node-count.sh {{.CLI_ARGS}}