kubeasz/ezctl

#!/bin/bash
#  Create & manage k8s clusters

set -o nounset
set -o errexit
#set -o xtrace

function usage() {
    echo -e "\033[33mUsage:\033[0m ezctl COMMAND [args]"
    cat <<EOF
-------------------------------------------------------------------------------------
Cluster setups:
    list		             to list all of the managed clusters
    checkout    <cluster>            to switch default kubeconfig of the cluster
    new         <cluster>            to start a new k8s deploy with name 'cluster'
    setup       <cluster>  <step>    to setup a cluster, also supporting a step-by-step way
    start       <cluster>            to start all of the k8s services stopped by 'ezctl stop'
    stop        <cluster>            to stop all of the k8s services temporarily
    upgrade     <cluster>            to upgrade the k8s cluster
    destroy     <cluster>            to destroy the k8s cluster
    backup      <cluster>            to backup the cluster state (etcd snapshot)
    restore     <cluster>            to restore the cluster state from backups
    start-aio		             to quickly setup an all-in-one cluster with 'default' settings

Cluster ops:
    add-etcd    <cluster>  <ip>      to add a etcd-node to the etcd cluster
    add-master  <cluster>  <ip>      to add a master node to the k8s cluster
    add-node    <cluster>  <ip>      to add a work node to the k8s cluster
    del-etcd    <cluster>  <ip>      to delete a etcd-node from the etcd cluster
    del-master  <cluster>  <ip>      to delete a master node from the k8s cluster
    del-node    <cluster>  <ip>      to delete a work node from the k8s cluster

Use "ezctl help <command>" for more information about a given command.
EOF
}

function usage-setup(){
  echo -e "\033[33mUsage:\033[0m ezctl setup <cluster> <step>"
  cat <<EOF
available steps:
    01     prepare         to prepare CA/certs & kubeconfig & other system settings 
    02     etcd            to setup the etcd cluster
    03     runtime         to setup the container runtime(docker or containerd)
    04     kube-master     to setup the master nodes
    05     kube-node       to setup the worker nodes
    06     network         to setup the network plugin
    07     cluster-addon   to setup other useful plugins
    all                    to run 01~07 all at once

examples: ./ezctl setup test-k8s 01
          ./ezctl setup test-k8s 02
          ./ezctl setup test-k8s all
EOF
}

function logger() {
  TIMESTAMP=$(date +'%Y-%m-%d %H:%M:%S')
  case "$1" in
    debug)
      echo -e "$TIMESTAMP \033[36mDEBUG\033[0m $2"
      ;;
    info)
      echo -e "$TIMESTAMP \033[32mINFO\033[0m $2"
      ;;
    warn)
      echo -e "$TIMESTAMP \033[33mWARN\033[0m $2"
      ;;
    error)
      echo -e "$TIMESTAMP \033[31mERROR\033[0m $2"
      ;;
    *)
      ;;
  esac
}

function help-info() {
    case "$1" in
        (setup)
	    usage-setup
            ;;
        (add-etcd)
            echo -e "read more > 'https://github.com/easzlab/kubeasz/blob/master/docs/op/op-etcd.md'"
            ;;
        (add-master)
            echo -e "read more > 'https://github.com/easzlab/kubeasz/blob/master/docs/op/op-master.md'"
            ;;
        (add-node)
            echo -e "read more > 'https://github.com/easzlab/kubeasz/blob/master/docs/op/op-node.md'"
            ;;
        (del-etcd)
            echo -e "read more > 'https://github.com/easzlab/kubeasz/blob/master/docs/op/op-etcd.md'"
            ;;
        (del-master)
            echo -e "read more > 'https://github.com/easzlab/kubeasz/blob/master/docs/op/op-master.md'"
            ;;
        (del-node)
            echo -e "read more > 'https://github.com/easzlab/kubeasz/blob/master/docs/op/op-node.md'"
            ;;
        (*)
            echo -e "todo: help info $1"
            ;;
    esac
}

### Cluster setups functions ##############################

function new() {
    # check if already existed
    [[ -d "clusters/$1" ]] && { logger error "cluster:$1 already existed, if cluster:$1 setup failed, try 'rm -rf clusters/$1' first!"; exit 1; }

    logger debug "generate custom cluster files in clusters/$1"
    mkdir -p "clusters/$1"
    cp example/hosts.multi-node "clusters/$1/hosts"
    sed -i "s/_cluster_name_/$1/g" "clusters/$1/hosts"
    cp example/config.yml "clusters/$1/config.yml"

    logger debug "set version of common plugins"
    calicoVer=$(grep 'calicoVer=' ezdown|cut -d'=' -f2)
    ciliumVer=$(grep 'ciliumVer=' ezdown|cut -d'=' -f2)
    flannelVer=$(grep 'flannelVer=' ezdown|cut -d'=' -f2)
    kubeRouterVer=$(grep 'kubeRouterVer=' ezdown|cut -d'=' -f2)
    kubeOvnVer=$(grep 'kubeOvnVer=' ezdown|cut -d'=' -f2)
    corednsVer=$(grep 'corednsVer=' ezdown|cut -d'=' -f2)
    dashboardVer=$(grep 'dashboardVer=' ezdown|cut -d'=' -f2)
    dashboardMetricsScraperVer=$(grep 'dashboardMetricsScraperVer=' ezdown|cut -d'=' -f2)
    metricsVer=$(grep 'metricsVer=' ezdown|cut -d'=' -f2)
    promChartVer=$(grep 'promChartVer=' ezdown|cut -d'=' -f2)
    traefikChartVer=$(grep 'traefikChartVer=' ezdown|cut -d'=' -f2)
    registryMirror=true

    grep registry-mirrors /etc/docker/daemon.json > /dev/null 2>&1 || { logger debug "disable registry mirrors"; registryMirror=false; }

    sed -i -e "s/__flannel__/$flannelVer/g" \
	   -e "s/__calico__/$calicoVer/g" \
	   -e "s/__cilium__/$ciliumVer/g" \
	   -e "s/__kube_ovn__/$kubeOvnVer/g" \
	   -e "s/__kube_router__/$kubeRouterVer/g" \
	   -e "s/__coredns__/$corednsVer/g" \
	   -e "s/__dashboard__/$dashboardVer/g" \
	   -e "s/__dash_metrics__/$dashboardMetricsScraperVer/g" \
	   -e "s/__prom_chart__/$promChartVer/g" \
	   -e "s/__traefik_chart__/$traefikChartVer/g" \
	   -e "s/^ENABLE_MIRROR_REGISTRY.*$/ENABLE_MIRROR_REGISTRY: $registryMirror/g" \
	   -e "s/__metrics__/$metricsVer/g" "clusters/$1/config.yml"


    logger debug "cluster $1: files successfully created."
    logger info "next steps 1: to config 'clusters/$1/hosts'"
    logger info "next steps 2: to config 'clusters/$1/config.yml'"
}

function setup() {
    [[ -d "clusters/$1" ]] || { logger error "invalid config, run 'ezctl new $1' first"; return 1; }
    [[ -f "bin/kube-apiserver" ]] || { logger error "no binaries founded, run 'ezdown -D' fist"; return 1; }

    PLAY_BOOK="dummy.yml"
    case "$2" in
      (01)
          PLAY_BOOK="01.prepare.yml"
          ;;
      (02)
          PLAY_BOOK="02.etcd.yml"
          ;;
      (03)
          PLAY_BOOK="03.runtime.yml"
          ;;
      (04)
          PLAY_BOOK="04.kube-master.yml"
          ;;
      (05)
          PLAY_BOOK="05.kube-node.yml"
          ;;
      (06)
          PLAY_BOOK="06.network.yml"
          ;;
      (07)
          PLAY_BOOK="07.cluster-addon.yml"
          ;;
      (90)
          PLAY_BOOK="90.setup.yml"
          ;;
      (all)
          PLAY_BOOK="90.setup.yml"
          ;;
      (*)
          usage-setup
          exit 1 
          ;;
    esac

    logger info "cluster:$1 setup step:$2 begins in 5s, press any key to abort:\n"
    ! (read -r -t5 -n1) || { logger warn "setup abort"; return 1; }

    ansible-playbook -i "clusters/$1/hosts" -e "@clusters/$1/config.yml" "playbooks/$PLAY_BOOK" || return 1
}

function cmd() {
    [[ -d "clusters/$1" ]] || { logger error "invalid config, run 'ezctl new $1' first"; return 1; }

    PLAY_BOOK="dummy.yml"
    case "$2" in
      (start)
          PLAY_BOOK="91.start.yml"
          ;;
      (stop)
          PLAY_BOOK="92.stop.yml"
          ;;
      (upgrade)
          PLAY_BOOK="93.upgrade.yml"
          ;;
      (backup)
          PLAY_BOOK="94.backup.yml"
          ;;
      (restore)
          PLAY_BOOK="95.restore.yml"
          ;;
      (destroy)
          PLAY_BOOK="99.clean.yml"
          ;;
      (*)
          usage
          exit 1
          ;;
    esac

    logger info "cluster:$1 $2 begins in 5s, press any key to abort:\n"
    ! (read -r -t5 -n1) || { logger warn "$2 abort"; return 1; }

    ansible-playbook -i "clusters/$1/hosts" -e "@clusters/$1/config.yml" "playbooks/$PLAY_BOOK" || return 1
}


function list() {
    [[ -d ./clusters ]] || { logger error "cluster not found, run 'ezctl new' first"; return 1; }
    [[ -f ~/.kube/config ]] || { logger error "kubeconfig not found, run 'ezctl setup' first"; return 1; }
    which md5sum > /dev/null 2>&1 || { logger error "md5sum not found"; return 1; }

    CLUSTERS=$(cd clusters && echo -- *)
    CFG_MD5=$(md5sum -t ~/.kube/config |cut -d' ' -f1)
    cd "$BASE"

    logger info "list of managed clusters:"
    i=1; for c in $CLUSTERS;
    do
        if [[ -f "clusters/$c/kubectl.kubeconfig" ]];then
            c_md5=$(md5sum -t "clusters/$c/kubectl.kubeconfig" |cut -d' ' -f1)
            if [[ "$c_md5" = "$CFG_MD5" ]];then
                echo -e "==> cluster $i:\t$c (\033[32mcurrent\033[0m)"
            else
                echo -e "==> cluster $i:\t$c"
            fi
            let "i++"
        fi
    done
}


function checkout() {
    [[ -d "clusters/$1" ]] || { logger error "invalid config, run 'ezctl new $1' first"; return 1; }
    [[ -f "clusters/$1/kubectl.kubeconfig" ]] || { logger error "invalid kubeconfig, run 'ezctl setup $1' first"; return 1; }
    logger info "set default kubeconfig: cluster $1 (\033[32mcurrent\033[0m)"
    /bin/cp -f "clusters/$1/kubectl.kubeconfig" ~/.kube/config
}

### in-cluster operation functions ##############################

function add-node() {
    # check new node's address regexp
    [[ $2 =~ ^(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})(\.(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})){3}$ ]] || { logger error "Invalid ip add:$2"; return 1; }

    # check if the new node already exsited
    sed -n '/^\[kube-master/,/^\[harbor/p' "$BASE/clusters/$1/hosts"|grep "^$2[^0-9]*$" && { logger error "node $2 already existed in $BASE/clusters/$1/hosts"; return 2; }

    logger info "add $2 into 'kube-node' group"
    sed -i "/\[kube-node/a $2 NEW_NODE=yes ${@:3}" "$BASE/clusters/$1/hosts"

    logger info "start to add a work node:$2 into cluster:$1"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/22.addnode.yml" -e "NODE_TO_ADD=$2" -e "@clusters/$1/config.yml"
}

function add-master() {
    # check new master's address regexp
    [[ $2 =~ ^(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})(\.(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})){3}$ ]] || { logger error  "Invalid ip add:$2"; return 1; }

    # check if the new master already exsited
    sed -n '/^\[kube-master/,/^\[kube-node/p' "$BASE/clusters/$1/hosts"|grep "^$2[^0-9]*$" && { logger error "master $2 already existed!"; return 2; }

    logger info "add $2 into 'kube-master' group"
    sed -i "/\[kube-master/a $2 NEW_MASTER=yes ${@:3}" "$BASE/clusters/$1/hosts"

    logger info "start to add a master node:$2 into cluster:$1"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/23.addmaster.yml" -e "NODE_TO_ADD=$2" -e "@clusters/$1/config.yml"

    logger info "reconfigure and restart the haproxy service on 'kube-node' nodes"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/05.kube-node.yml" -t restart_lb -e MASTER_CHG=yes -e "@clusters/$1/config.yml" 
}

function add-etcd() {
    # check new node's address regexp
    [[ $2 =~ ^(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})(\.(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})){3}$ ]] || { logger error  "Invalid ip add:$2"; return 1; }

    # check if the new node already exsited
    sed -n '/^\[etcd/,/^\[kube-master/p' "$BASE/clusters/$1/hosts"|grep "^$2[^0-9]*$" && { logger error "etcd $2 already existed!"; return 2; }

    logger info "add $2 into 'etcd' group"
    sed -i "/\[etcd/a $2 NEW_ETCD=yes ${@:3}" "$BASE/clusters/$1/hosts" 

    logger info "start to add a etcd node:$2 into cluster:$1"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/21.addetcd.yml" -e "NODE_TO_ADD=$2" -e "@clusters/$1/config.yml"

    logger info "reconfig &restart the etcd cluster"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/02.etcd.yml" -t restart_etcd -e "@clusters/$1/config.yml"
    
    logger info "restart apiservers to use the new etcd cluster"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/04.kube-master.yml" -t restart_master -e "@clusters/$1/config.yml" 
}

function del-etcd() {
    # check node's address regexp
    [[ $2 =~ ^(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})(\.(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})){3}$ ]] || { logger error  "Invalid ip add:$2"; return 1; }

    logger warn "start to delete the etcd node:$2 from cluster:$1"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/31.deletcd.yml" -e "ETCD_TO_DEL=$2" -e "CLUSTER=$1" -e "@clusters/$1/config.yml"

    logger info "reconfig &restart the etcd cluster"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/02.etcd.yml" -t restart_etcd -e "@clusters/$1/config.yml"
    
    logger info "restart apiservers to use the new etcd cluster"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/04.kube-master.yml" -t restart_master -e "@clusters/$1/config.yml" 
}

function del-node() {
    # check node's address regexp
    [[ $2 =~ ^(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})(\.(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})){3}$ ]] || { logger "Invalid ip add:$2"; return 2; }

    logger warn "start to delete the node:$2 from cluster:$1"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/32.delnode.yml" -e "NODE_TO_DEL=$2" -e "CLUSTER=$1" -e "@clusters/$1/config.yml"
}

function del-master() {
    # check node's address regexp
    [[ $2 =~ ^(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})(\.(2(5[0-5]{1}|[0-4][0-9]{1})|[0-1]?[0-9]{1,2})){3}$ ]] || { logger error "Invalid ip add:$2"; return 2; }

    logger warn "start to delete the master:$2 from cluster:$1"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/33.delmaster.yml" -e "NODE_TO_DEL=$2" -e "CLUSTER=$1" -e "@clusters/$1/config.yml" 

    logger info "reconfig kubeconfig in ansible manage node"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/roles/deploy/deploy.yml" -t create_kctl_cfg -e "@clusters/$1/config.yml"

    logger info "reconfigure and restart the haproxy service on 'kube-node' nodes"
    ansible-playbook -i "$BASE/clusters/$1/hosts" "$BASE/playbooks/05.kube-node.yml" -t restart_lb -e MASTER_CHG=yes -e "@clusters/$1/config.yml" 
}


function start-aio(){
    set +u
    # Check ENV 'HOST_IP', exists if the CMD 'ezctl' running in a docker container
    if [[ -z $HOST_IP ]];then
        # ezctl runs in a host machine, get host's ip
        HOST_IF=$(ip route|grep default|cut -d' ' -f5)
        HOST_IP=$(ip a|grep "$HOST_IF$"|head -n1|awk '{print $2}'|cut -d'/' -f1)
    fi
    set -u

    logger info "get local host ipadd: $HOST_IP"

    new default
    /bin/cp -f example/hosts.allinone "clusters/default/hosts"
    sed -i "s/_cluster_name_/default/g" "clusters/default/hosts"
    sed -i "s/192.168.1.1/$HOST_IP/g" "clusters/default/hosts"

    setup default all
}


### Main Lines ##################################################
function main() {
  BASE="/etc/kubeasz"
  [[ -d "$BASE" ]] || { logger error "invalid dir:$BASE, try: 'ezdown -D'"; exit 1; }
  cd "$BASE"

  # check bash shell
  readlink /proc/$$/exe|grep -q "dash" && { logger error "you should use bash shell only"; exit 1; }

  # check 'ansible' executable
  which ansible > /dev/null 2>&1 || { logger error "need 'ansible', try: 'pip install ansible==2.6.18'"; usage; exit 1; }
  
  [ "$#" -gt 0 ] || { usage >&2; exit 2; }
  
  case "$1" in
      ### in-cluster operations #####################
      (add-etcd)
          [ "$#" -gt 2 ] || { usage >&2; exit 2; }
          add-etcd "${@:2}"
          ;;
      (add-master)
          [ "$#" -gt 2 ] || { usage >&2; exit 2; }
          add-master "${@:2}"
          ;;
      (add-node)
          [ "$#" -gt 2 ] || { usage >&2; exit 2; }
          add-node "${@:2}"
          ;;
      (del-etcd)
          [ "$#" -eq 3 ] || { usage >&2; exit 2; }
          del-etcd "$2" "$3"
          ;;
      (del-master)
          [ "$#" -eq 3 ] || { usage >&2; exit 2; }
          del-master "$2" "$3"
          ;;
      (del-node)
          [ "$#" -eq 3 ] || { usage >&2; exit 2; }
          del-node "$2" "$3"
          ;;
      ### cluster-wide operations #######################
      (checkout)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          checkout "$2"
          ;;
      (list)
          [ "$#" -eq 1 ] || { usage >&2; exit 2; }
          list
          ;;
      (new)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          new "$2"
          ;;
      (setup)
          [ "$#" -eq 3 ] || { usage-setup >&2; exit 2; }
          setup "${@:2}"
          ;;
      (start)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          cmd "$2" start
          ;;
      (stop)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          cmd "$2" stop
          ;;
      (upgrade)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          cmd "$2" upgrade
          ;;
      (backup)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          cmd "$2" backup
          ;;
      (restore)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          cmd "$2" restore
          ;;
      (destroy)
          [ "$#" -eq 2 ] || { usage >&2; exit 2; }
          cmd "$2" destroy
          ;;
      (start-aio)
          [ "$#" -eq 1 ] || { usage >&2; exit 2; }
          start-aio
          ;;
      (help)
          [ "$#" -gt 1 ] || { usage >&2; exit 2; }
          help-info "$2"
          exit 0
          ;;
      (*)
          usage
          exit 0
          ;;
  esac
 }

main "$@"