# Ignore SSH key validation when connecting to these remote hosts.
# (Otherwise, deployment scripts break when a VM IP address reuse.)
SSHOPTS="-o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -o LogLevel=ERROR"

HELP=""
_cmd() {
    HELP="$(printf "%s\n%-20s %s\n" "$HELP" "$1" "$2")"
}

_cmd help "Show available commands"
_cmd_help() {
    printf "$(basename $0) - the container training swiss army knife\n"
    printf "Commands:"
    printf "%s" "$HELP" | sort
}

_cmd cards "Generate ready-to-print cards for a group of VMs"
_cmd_cards() {
    TAG=$1
    need_tag

    OPTIONS_FILE=$2
    [ -f "$OPTIONS_FILE" ] || die "Please specify a YAML options file as 2nd argument."
    OPTIONS_FILE_PATH="$(readlink -f "$OPTIONS_FILE")"

    # This will process logins.jsonl to generate two files: cards.pdf and cards.html
    (
        cd tags/$TAG
        ../../../lib/make-login-cards.py "$OPTIONS_FILE_PATH"
    )

    ln -sf ../tags/$TAG/cards.html www/$TAG.html
    ln -sf ../tags/$TAG/cards.pdf www/$TAG.pdf

    info "Cards created. You can view them with:"
    info "xdg-open tags/$TAG/cards.html tags/$TAG/cards.pdf (on Linux)"
    info "open tags/$TAG/cards.html (on macOS)"
    info "Or you can start a web server with:"
    info "$0 www"
}

_cmd clean "Remove information about destroyed clusters"
_cmd_clean() {
	for TAG in tags/*; do
		if grep -q ^destroyed$ "$TAG/status"; then
			info "Removing $TAG..."
			rm -rf "$TAG"
		fi
	done	
}

_cmd codeserver "Install code-server on the clusters"
_cmd_codeserver() {
    TAG=$1
    need_tag

    ARCH=${ARCHITECTURE-amd64}
    CODESERVER_VERSION=4.96.4
    CODESERVER_URL=https://github.com/coder/code-server/releases/download/v${CODESERVER_VERSION}/code-server-${CODESERVER_VERSION}-linux-${ARCH}.tar.gz
    pssh "
    set -e
    i_am_first_node || exit 0
    if ! [ -x /usr/local/bin/code-server ]; then
        curl -fsSL $CODESERVER_URL | sudo tar zx -C /opt
        sudo ln -s /opt/code-server-${CODESERVER_VERSION}-linux-${ARCH}/bin/code-server /usr/local/bin/code-server
        sudo -u $USER_LOGIN -H code-server --install-extension ms-azuretools.vscode-docker
        sudo -u $USER_LOGIN -H code-server --install-extension ms-kubernetes-tools.vscode-kubernetes-tools
        sudo -u $USER_LOGIN -H mkdir -p /home/$USER_LOGIN/.local/share/code-server/User
        echo '{\"workbench.startupEditor\": \"terminal\"}' | sudo -u $USER_LOGIN tee /home/$USER_LOGIN/.local/share/code-server/User/settings.json
        sudo -u $USER_LOGIN mkdir -p /home/$USER_LOGIN/.config/systemd/user
        sudo -u $USER_LOGIN tee /home/$USER_LOGIN/.config/systemd/user/code-server.service <<EOF
[Unit]
Description=code-server

[Install]
WantedBy=default.target

[Service]
ExecStart=/usr/local/bin/code-server --bind-addr [::]:1789
Restart=always
EOF
        sudo systemctl --user -M $USER_LOGIN@ enable code-server.service --now
        sudo loginctl enable-linger $USER_LOGIN
    fi"
}

_cmd createuser "Create the user that students will use"
_cmd_createuser() {
    TAG=$1
    need_tag

    pssh "
    set -e
    # Create the user if it doesn't exist yet.
    id $USER_LOGIN || sudo useradd -d /home/$USER_LOGIN -g users -m -s /bin/bash $USER_LOGIN
    # Make sure there are at least exec permission on their home.
    sudo chmod a+X /home/$USER_LOGIN
    # Add them to the docker group, if there is one.
    grep ^docker: /etc/group && sudo usermod -aG docker $USER_LOGIN
    # Set their password.
    echo $USER_LOGIN:$USER_PASSWORD | sudo chpasswd
    # Add them to sudoers and allow passwordless authentication.
    echo '$USER_LOGIN ALL=(ALL) NOPASSWD:ALL' | sudo tee /etc/sudoers.d/$USER_LOGIN
    "

    # The MaxAuthTries is here to help with folks who have many SSH keys.
    pssh "
    set -e
    sudo sed -i 's/PasswordAuthentication no/PasswordAuthentication yes/' /etc/ssh/sshd_config
    sudo sed -i 's/#MaxAuthTries 6/MaxAuthTries 42/' /etc/ssh/sshd_config
    sudo systemctl restart ssh.service
    "

    pssh "
    set -e
    cd /home/$USER_LOGIN
    sudo -u $USER_LOGIN mkdir -p .ssh
    if i_am_first_node; then
      # Generate a key pair with an empty passphrase.
      if ! sudo -u $USER_LOGIN [ -f .ssh/id_rsa ]; then
        sudo -u $USER_LOGIN ssh-keygen -t rsa -f .ssh/id_rsa -P ''
        sudo -u $USER_LOGIN cp .ssh/id_rsa.pub .ssh/authorized_keys
      fi
    fi
    "

    # FIXME this is a gross hack to add the deployment key to our SSH agent,
    # so that it can be used to bounce from host to host (which is necessary
    # in the next deployment step). In the long run, we probably want to
    # generate these keys locally and push them to the machines instead
    # (once we move everything to Terraform).
    ssh-add tags/$TAG/id_rsa
    pssh "
    set -e
    cd /home/$USER_LOGIN
    if ! i_am_first_node; then
      # Copy keys from the first node.
      ssh $SSHOPTS \$(cat /etc/name_of_first_node) sudo -u $USER_LOGIN tar -C /home/$USER_LOGIN -cvf- .ssh |
      sudo -u $USER_LOGIN tar -xf-
    fi
    "
    ssh-add -d tags/$TAG/id_rsa

    # FIXME do this only once.
    pssh -I "sudo -u $USER_LOGIN tee -a /home/$USER_LOGIN/.bashrc" <<"SQRL"

# Fancy prompt courtesy of @soulshake.
export PS1='\e[1m\e[31m[$HOSTIP] \e[32m($(docker-prompt)) \e[34m\u@\h\e[35m \w\e[0m\n$ '

# Bigger history, in a different file, and saved before executing each command.
export HISTSIZE=9999
export HISTFILESIZE=9999
shopt -s histappend
trap 'history -a' DEBUG
export HISTFILE=~/.history
SQRL

    pssh -I "sudo -u $USER_LOGIN tee /home/$USER_LOGIN/.vimrc" <<SQRL
syntax on
set autoindent
set expandtab
set number
set shiftwidth=2
set softtabstop=2
set nowrap
set laststatus=2
SQRL

    pssh -I "sudo -u $USER_LOGIN tee /home/$USER_LOGIN/.tmux.conf" <<SQRL
set -g status-style bg=yellow,bold

bind h select-pane -L
bind j select-pane -D
bind k select-pane -U
bind l select-pane -R

# Allow using mouse to switch panes
set -g mouse on

# Make scrolling with wheels work
bind -n WheelUpPane if-shell -F -t = "#{mouse_any_flag}" "send-keys -M" "if -Ft= '#{pane_in_mode}' 'send-keys -M' 'select-pane -t=; copy-mode -e; send-keys -M'"
bind -n WheelDownPane select-pane -t= \; send-keys -M

# Retain one million lines
set-option -g history-limit 1000000
SQRL

    # Install docker-prompt script
    pssh -I sudo tee /usr/local/bin/docker-prompt <lib/docker-prompt
    pssh sudo chmod +x /usr/local/bin/docker-prompt

    echo user_ok > tags/$TAG/status
}


_cmd create "Create lab environments"
_cmd_create() {
    while [ ! -z "$*" ]; do
        case "$1" in
        --mode) MODE=$2; shift 2;;
        --provider) PROVIDER=$2; shift 2;;
        --settings) SETTINGS=$2; shift 2;;
        --students) STUDENTS=$2; shift 2;;
        --tag) TAG=$2; shift 2;;
        *) die "Unrecognized parameter: $1."
        esac
    done

    if [ -z "$MODE" ]; then
        info "Using default mode (pssh)."
        MODE=pssh
    fi
    if [ -z "$PROVIDER" ]; then
        die "Please add --provider flag to specify which provider to use."
    fi
    if [ -z "$SETTINGS" ]; then
        die "Please add --settings flag to specify which settings file to use."
    fi
    if [ -z "$STUDENTS" ]; then
        info "Defaulting to 1 student since --students flag wasn't specified."
        STUDENTS=1
    fi

    case "$MODE" in
        mk8s)
            PROVIDER_BASE=terraform/one-kubernetes
            ;;
        pssh)
            PROVIDER_BASE=terraform/virtual-machines
            ;;
        *) die "Invalid mode: $MODE (supported modes: mk8s, pssh)." ;;
    esac
    
    if ! [ -f "$SETTINGS" ]; then
        die "Settings file ($SETTINGS) not found."
    fi

    # Check that the provider is valid.
    if [ -d $PROVIDER_BASE/$PROVIDER ]; then
        if [ -f $PROVIDER_BASE/$PROVIDER/requires_tfvars ]; then
            die "Provider $PROVIDER cannot be used directly, because it requires a tfvars file."
        fi
        PROVIDER_DIRECTORY=$PROVIDER_BASE/$PROVIDER
        TFVARS=""
    elif [ -f $PROVIDER_BASE/$PROVIDER.tfvars ]; then
        TFVARS=$PROVIDER_BASE/$PROVIDER.tfvars
        PROVIDER_DIRECTORY=$(dirname $PROVIDER_BASE/$PROVIDER)
    else
        error "Provider $PROVIDER not found."
        info "Available providers for mode $MODE:"
        (
            cd $PROVIDER_BASE
            for P in *; do
                if [ -d "$P" ]; then
                    [ -f "$P/requires_tfvars" ] || info "$P"
                    for V in $P/*.tfvars; do
                        [ -f "$V" ] && info "${V%.tfvars}"
                    done
                fi
            done
        )
        die "Please specify a valid provider."
    fi

    if [ -z "$TAG" ]; then
        TAG=$(_cmd_maketag)
    fi
    mkdir -p tags/$TAG
    echo creating > tags/$TAG/status

    ln -s ../../$SETTINGS tags/$TAG/settings.env.orig
    cp $SETTINGS tags/$TAG/settings.env
    . $SETTINGS

    echo $MODE > tags/$TAG/mode
    echo $PROVIDER > tags/$TAG/provider
    case "$MODE" in
        mk8s)
            cp -d terraform/many-kubernetes/*.* tags/$TAG
            mkdir tags/$TAG/one-kubernetes-module
            cp $PROVIDER_DIRECTORY/*.tf tags/$TAG/one-kubernetes-module
            mkdir tags/$TAG/one-kubernetes-config
            mv tags/$TAG/one-kubernetes-module/config.tf tags/$TAG/one-kubernetes-config
            ;;
        pssh)
            cp $PROVIDER_DIRECTORY/*.tf tags/$TAG
            if [ "$TFVARS" ]; then
                cp "$TFVARS" "tags/$TAG/$(basename $TFVARS).auto.tfvars"
            fi
            ;;
    esac
    (
        cd tags/$TAG
        terraform init
        echo tag = \"$TAG\" >> terraform.tfvars
        echo how_many_clusters = $STUDENTS >> terraform.tfvars
        if [ "$CLUSTERSIZE" ]; then
            echo nodes_per_cluster = $CLUSTERSIZE >> terraform.tfvars
        fi
    )

    sep

    # If the settings.env file has a "STEPS" field,
    # automatically execute all the actions listed in that field.
    # If an action fails, retry it up to 10 times.
    for STEP in $(echo $STEPS); do
        sep "$TAG -> $STEP"
        TRY=1
        MAXTRY=10
        while ! $0 $STEP $TAG ; do
            TRY=$(($TRY+1))
            if [ $TRY -gt $MAXTRY ]; then
                error "This step ($STEP) failed after $MAXTRY attempts."
                info "You can troubleshoot the situation manually, or terminate these instances with:"
                info "$0 destroy $TAG"
                die "Giving up."
            else
                sep
                info "Step '$STEP' failed for '$TAG'. Let's wait 10 seconds and try again."
                info "(Attempt $TRY out of $MAXTRY.)"
                sleep 10
            fi
        done
    done
    sep
    info "Deployment successful."
    info "To log into the first machine of that batch, you can run:"
    info "$0 ssh $TAG"
    info "To terminate these instances, you can run:"
    info "$0 destroy $TAG"
}

_cmd destroy "Destroy lab environments"
_cmd_destroy() {
    TAG=$1
    need_tag
    cd tags/$TAG
    echo destroying > status
    terraform destroy -auto-approve
    echo destroyed > status
}

_cmd clusterize "Group VMs in clusters"
_cmd_clusterize() {
    TAG=$1
    need_tag

    pssh "
    set -e
    grep PSSH_ /etc/ssh/sshd_config || echo 'AcceptEnv PSSH_*' | sudo tee -a /etc/ssh/sshd_config
    grep KUBECOLOR_ /etc/ssh/sshd_config || echo 'AcceptEnv KUBECOLOR_*' | sudo tee -a /etc/ssh/sshd_config
    sudo systemctl restart ssh.service"

    pssh -I < tags/$TAG/clusters.tsv "
    grep -w \$PSSH_HOST | tr '\t' '\n' > /tmp/cluster"
    pssh "
    echo \$PSSH_HOST > /tmp/ipv4
    head -n 1 /tmp/cluster | sudo tee /etc/ipv4_of_first_node
    echo ${CLUSTERPREFIX}1 | sudo tee /etc/name_of_first_node
    echo HOSTIP=\$PSSH_HOST | sudo tee -a /etc/environment
    NODEINDEX=\$((\$PSSH_NODENUM%$CLUSTERSIZE+1))
    if [ \$NODEINDEX = 1 ]; then
        sudo ln -sf /bin/true /usr/local/bin/i_am_first_node
    else
        sudo ln -sf /bin/false /usr/local/bin/i_am_first_node
    fi
    echo $CLUSTERPREFIX\$NODEINDEX | sudo tee /etc/hostname
    sudo hostname $CLUSTERPREFIX\$NODEINDEX
    N=1
    while read ip; do
        grep -w \$ip /etc/hosts || echo \$ip $CLUSTERPREFIX\$N | sudo tee -a /etc/hosts
        N=\$((\$N+1))
    done < /tmp/cluster
    "

    jq --raw-input --compact-output \
       --arg USER_LOGIN "$USER_LOGIN" --arg USER_PASSWORD "$USER_PASSWORD" '
    {
      "login": $USER_LOGIN,
      "password": $USER_PASSWORD,
      "ipaddrs": .
    }' < tags/$TAG/clusters.tsv > tags/$TAG/logins.jsonl

    echo cluster_ok > tags/$TAG/status
}

_cmd disabledocker "Stop Docker Engine and don't restart it automatically"
_cmd_disabledocker() {
    TAG=$1
    need_tag

    pssh "
    sudo systemctl disable docker.socket --now
    sudo systemctl disable docker.service --now
    sudo systemctl disable containerd.service --now
    "
}

_cmd docker "Install and start Docker"
_cmd_docker() {
    TAG=$1
    need_tag

    pssh "
    set -e
    # On EC2, the ephemeral disk might be mounted on /mnt.
    # If /mnt is a mountpoint, place Docker workspace on it.
    if mountpoint -q /mnt; then
      sudo mkdir -p /mnt/docker
      sudo ln -sfn /mnt/docker /var/lib/docker
    fi

    # This will install the latest Docker.
    sudo apt-get -qy install apt-transport-https ca-certificates curl software-properties-common
    curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
    sudo add-apt-repository 'deb https://download.docker.com/linux/ubuntu jammy stable'
    sudo apt-get -q update
    sudo apt-get -qy install docker-ce

    # Add registry mirror configuration.
    if ! [ -f /etc/docker/daemon.json ]; then
        sudo mkdir -p /etc/docker
        echo '{\"registry-mirrors\": [\"https://mirror.gcr.io\"]}' | sudo tee /etc/docker/daemon.json
        sudo systemctl restart docker
    fi
    "

    ##VERSION## https://github.com/docker/compose/releases
    COMPOSE_VERSION=v2.11.1
    COMPOSE_PLATFORM='linux-$(uname -m)'

    # Just in case you need Compose 1.X, you can use the following lines.
    # (But it will probably only work for x86_64 machines.)
    #COMPOSE_VERSION=1.29.2
    #COMPOSE_PLATFORM='Linux-$(uname -m)'

    pssh "
    set -e
    ### Install docker-compose.
    sudo curl -fsSL -o /usr/local/bin/docker-compose \
      https://github.com/docker/compose/releases/download/$COMPOSE_VERSION/docker-compose-$COMPOSE_PLATFORM
    sudo chmod +x /usr/local/bin/docker-compose
    docker-compose version

    ### Install docker-machine.
    ##VERSION## https://github.com/docker/machine/releases
    MACHINE_VERSION=v0.16.2
    sudo curl -fsSL -o /usr/local/bin/docker-machine \
      https://github.com/docker/machine/releases/download/\$MACHINE_VERSION/docker-machine-\$(uname -s)-\$(uname -m)
    sudo chmod +x /usr/local/bin/docker-machine
    docker-machine version
    "
}

_cmd kubebins "Install Kubernetes and CNI binaries but don't start anything"
_cmd_kubebins() {
    TAG=$1
    need_tag

    if [ "$KUBEVERSION" = "" ]; then
        KUBEVERSION="$(curl -fsSL https://cdn.dl.k8s.io/release/stable.txt | sed s/^v//)"
    fi

    ##VERSION##
    case "$KUBEVERSION" in
    1.19.*)
      ETCD_VERSION=v3.4.13
      CNI_VERSION=v0.8.7
      ;;
    *)
      ETCD_VERSION=v3.5.10
      CNI_VERSION=v1.3.0
      ;;
    esac

    K8SBIN_VERSION="v$KUBEVERSION"
    ARCH=${ARCHITECTURE-amd64}
    pssh --timeout 300 "
    set -e
    cd /usr/local/bin
    if ! [ -x etcd ]; then
        curl -L https://github.com/etcd-io/etcd/releases/download/$ETCD_VERSION/etcd-$ETCD_VERSION-linux-$ARCH.tar.gz \
        | sudo tar --strip-components=1 --wildcards -zx '*/etcd' '*/etcdctl'
    fi
    if ! [ -x hyperkube ]; then
        ##VERSION##
        curl -L https://dl.k8s.io/$K8SBIN_VERSION/kubernetes-server-linux-$ARCH.tar.gz \
        | sudo tar --strip-components=3 -zx \
          kubernetes/server/bin/kube{ctl,let,-proxy,-apiserver,-scheduler,-controller-manager}
    fi
    sudo mkdir -p /opt/cni/bin
    cd /opt/cni/bin
    if ! [ -x bridge ]; then
        curl -L https://github.com/containernetworking/plugins/releases/download/$CNI_VERSION/cni-plugins-linux-$ARCH-$CNI_VERSION.tgz \
        | sudo tar -zx
    fi
    "
}

_cmd kubepkgs "Install Kubernetes packages (kubectl, kubeadm, kubelet)"
_cmd_kubepkgs() {
    TAG=$1
    need_tag

    # Prior September 2023, there was a single Kubernetes package repo that
    # contained packages for all versions, so we could just add that repo
    # and install whatever was the latest version available there.
    # Things have changed (versions after September 2023, e.g. 1.28.3 are
    # not in the old repo) and now there is a different repo for each
    # minor version, so we need to figure out what minor version we are
    # installing to add the corresponding repo.
    if [ "$KUBEVERSION" = "" ]; then
        KUBEVERSION="$(curl -fsSL https://cdn.dl.k8s.io/release/stable.txt | sed s/^v//)"
    fi
    KUBEREPOVERSION="$(echo $KUBEVERSION | cut -d. -f1-2)"

    # Since the new repo doesn't have older versions, add a safety check here.
    MINORVERSION="$(echo $KUBEVERSION | cut -d. -f2)"
    if [ "$MINORVERSION" -lt 24 ]; then
        die "Cannot install kubepkgs for versions before 1.24."
    fi

    pssh "
    sudo tee /etc/apt/preferences.d/kubernetes <<EOF
Package: kubectl kubeadm kubelet
Pin: version $KUBEVERSION-*
Pin-Priority: 1000
EOF"

    # Install packages
    pssh --timeout 200 "
    curl -fsSL https://pkgs.k8s.io/core:/stable:/v$KUBEREPOVERSION/deb/Release.key |
    gpg --dearmor | sudo tee /etc/apt/keyrings/kubernetes-apt-keyring.gpg &&
    echo 'deb [signed-by=/etc/apt/keyrings/kubernetes-apt-keyring.gpg] https://pkgs.k8s.io/core:/stable:/v$KUBEREPOVERSION/deb/ /' |
    sudo tee /etc/apt/sources.list.d/kubernetes.list"
    pssh --timeout 200 "
    sudo apt-get update -q &&
    sudo apt-get install -qy kubelet kubeadm kubectl &&
    sudo apt-mark hold kubelet kubeadm kubectl &&
    kubeadm completion bash | sudo tee /etc/bash_completion.d/kubeadm &&
    kubectl completion bash | sudo tee /etc/bash_completion.d/kubectl &&
    echo 'alias k=kubecolor' | sudo tee /etc/bash_completion.d/k &&
    echo 'complete -F __start_kubectl k' | sudo tee -a /etc/bash_completion.d/k"
}

_cmd kubeadm "Setup kubernetes clusters with kubeadm"
_cmd_kubeadm() {
    TAG=$1
    need_tag

    if [ "$KUBEVERSION" ]; then
        CLUSTER_CONFIGURATION_KUBERNETESVERSION='kubernetesVersion: "v'$KUBEVERSION'"'
        IGNORE_SYSTEMVERIFICATION="- SystemVerification"
        IGNORE_SWAP="- Swap"
        IGNORE_IPTABLES="- FileContent--proc-sys-net-bridge-bridge-nf-call-iptables"
    fi

    # Install a valid configuration for containerd
    # (first, the CRI interface needs to be re-enabled;
    # also, the correct systemd cgroup driver must be selected,
    # otherwise containerd just restarts containers for no good reason)
    pssh -I "sudo tee /etc/containerd/config.toml" < lib/containerd-config.toml
    pssh "sudo systemctl restart containerd"

    # Initialize kube control plane
    pssh --timeout 200 "
    if i_am_first_node && [ ! -f /etc/kubernetes/admin.conf ]; then
        kubeadm token generate > /tmp/token &&
        cat >/tmp/kubeadm-config.yaml <<EOF
kind: InitConfiguration
apiVersion: kubeadm.k8s.io/v1beta3
bootstrapTokens:
- token: \$(cat /tmp/token)
nodeRegistration:
  ignorePreflightErrors:
  - NumCPU
  $IGNORE_SYSTEMVERIFICATION
  $IGNORE_SWAP
  $IGNORE_IPTABLES
---
kind: JoinConfiguration
apiVersion: kubeadm.k8s.io/v1beta3
discovery:
  bootstrapToken:
    apiServerEndpoint: \$(cat /etc/name_of_first_node):6443
    token: \$(cat /tmp/token)
    unsafeSkipCAVerification: true
nodeRegistration:
  ignorePreflightErrors:
  - NumCPU
  $IGNORE_SYSTEMVERIFICATION
  $IGNORE_SWAP
  $IGNORE_IPTABLES
---
kind: KubeletConfiguration
apiVersion: kubelet.config.k8s.io/v1beta1
failSwapOn: false
---
kind: ClusterConfiguration
apiVersion: kubeadm.k8s.io/v1beta3
apiServer:
  certSANs:
  - \$(cat /tmp/ipv4)
$CLUSTER_CONFIGURATION_KUBERNETESVERSION
EOF
	sudo kubeadm init --config=/tmp/kubeadm-config.yaml
    fi"

    # Put kubeconfig in ubuntu's and $USER_LOGIN's accounts
    pssh "
    if i_am_first_node; then
        sudo mkdir -p \$HOME/.kube /home/$USER_LOGIN/.kube &&
        sudo cp /etc/kubernetes/admin.conf \$HOME/.kube/config &&
        sudo cp /etc/kubernetes/admin.conf /home/$USER_LOGIN/.kube/config &&
        sudo chown -R \$(id -u) \$HOME/.kube &&
        sudo chown -R $USER_LOGIN /home/$USER_LOGIN/.kube
    fi"

    # Install weave as the pod network
    pssh "
    if i_am_first_node; then
        curl -fsSL https://github.com/weaveworks/weave/releases/download/v2.8.1/weave-daemonset-k8s-1.11.yaml |
        sed s,weaveworks/weave,quay.io/rackspace/weave, |
        kubectl apply -f-
    fi"

    # FIXME this is a gross hack to add the deployment key to our SSH agent,
    # so that it can be used to bounce from host to host (which is necessary
    # in the next deployment step). In the long run, we probably want to
    # generate these keys locally and push them to the machines instead
    # (once we move everything to Terraform).
    if [ -f "tags/$TAG/id_rsa" ]; then
        ssh-add tags/$TAG/id_rsa
    fi
    # Join the other nodes to the cluster
    pssh --timeout 200 "
    if ! i_am_first_node && [ ! -f /etc/kubernetes/kubelet.conf ]; then
        FIRSTNODE=\$(cat /etc/name_of_first_node) &&
        ssh $SSHOPTS \$FIRSTNODE cat /tmp/kubeadm-config.yaml > /tmp/kubeadm-config.yaml &&
        sudo kubeadm join --config /tmp/kubeadm-config.yaml
    fi"
    if [ -f "tags/$TAG/id_rsa" ]; then
        ssh-add -d tags/$TAG/id_rsa
    fi

    # Install metrics server
    pssh "
    if i_am_first_node; then
	kubectl apply -f https://raw.githubusercontent.com/jpetazzo/container.training/master/k8s/metrics-server.yaml
    #helm upgrade --install metrics-server \
    #     --repo https://kubernetes-sigs.github.io/metrics-server/ metrics-server \
    #     --namespace kube-system --set args={--kubelet-insecure-tls}
    fi"
}

_cmd kubetools "Install a bunch of CLI tools for Kubernetes"
_cmd_kubetools() {
    TAG=$1
    need_tag

    ARCH=${ARCHITECTURE-amd64}

    # Folks, please, be consistent!
    # Either pick "uname -m" (on Linux, that's x86_64, aarch64, etc.)
    # Or GOARCH (amd64, arm64, etc.)
    # But don't mix both! Thank you ♥
    case $ARCH in
    amd64)
        HERP_DERP_ARCH=x86_64
        TILT_ARCH=x86_64
        ;;
    *)
        HERP_DERP_ARCH=$ARCH
        TILT_ARCH=${ARCH}_ALPHA
        ;;
    esac

    # Install ArgoCD CLI
    ##VERSION## https://github.com/argoproj/argo-cd/releases/latest
    URL=https://github.com/argoproj/argo-cd/releases/latest/download/argocd-linux-${ARCH}
    pssh "
    if [ ! -x /usr/local/bin/argocd ]; then
        sudo curl -o /usr/local/bin/argocd -fsSL $URL
        sudo chmod +x /usr/local/bin/argocd
        argocd completion bash | sudo tee /etc/bash_completion.d/argocd
        argocd version --client
    fi"

    # Install Flux CLI
    ##VERSION## https://github.com/fluxcd/flux2/releases
    FLUX_VERSION=2.3.0
    FILENAME=flux_${FLUX_VERSION}_linux_${ARCH}
    URL=https://github.com/fluxcd/flux2/releases/download/v$FLUX_VERSION/$FILENAME.tar.gz
    pssh "
    if [ ! -x /usr/local/bin/flux ]; then
        curl -fsSL $URL |
        sudo tar -C /usr/local/bin -zx flux
        sudo chmod +x /usr/local/bin/flux
        flux completion bash | sudo tee /etc/bash_completion.d/flux
        flux --version
    fi"

    # Install kubectx and kubens
    pssh "
    set -e
    if ! [ -x /usr/local/bin/kctx ]; then
      cd /tmp
      git clone https://github.com/ahmetb/kubectx
      sudo cp kubectx/kubectx /usr/local/bin/kctx
      sudo cp kubectx/kubens /usr/local/bin/kns
      sudo cp kubectx/completion/*.bash /etc/bash_completion.d
    fi"

    # Install kube-ps1
    pssh "
    set -e
    if ! [ -d /opt/kube-ps1 ]; then
      cd /tmp
      git clone https://github.com/jonmosco/kube-ps1
      sudo mv kube-ps1 /opt/kube-ps1
      sudo -u $USER_LOGIN sed -i s/docker-prompt/kube_ps1/ /home/$USER_LOGIN/.bashrc &&
      sudo -u $USER_LOGIN tee -a /home/$USER_LOGIN/.bashrc <<EOF
. /opt/kube-ps1/kube-ps1.sh
KUBE_PS1_PREFIX=""
KUBE_PS1_SUFFIX=""
KUBE_PS1_SYMBOL_ENABLE="false"
KUBE_PS1_CTX_COLOR="green"
KUBE_PS1_NS_COLOR="green"
EOF
    fi"

    # Install stern
    ##VERSION## https://github.com/stern/stern/releases
    STERN_VERSION=1.29.0
    FILENAME=stern_${STERN_VERSION}_linux_${ARCH}
    URL=https://github.com/stern/stern/releases/download/v$STERN_VERSION/$FILENAME.tar.gz
    pssh "
    if [ ! -x /usr/local/bin/stern ]; then
        curl -fsSL $URL |
        sudo tar -C /usr/local/bin -zx stern
        sudo chmod +x /usr/local/bin/stern
        stern --completion bash | sudo tee /etc/bash_completion.d/stern
        stern --version
    fi"

    # Install helm
    pssh "
    if [ ! -x /usr/local/bin/helm ]; then
        curl https://raw.githubusercontent.com/kubernetes/helm/master/scripts/get-helm-3 | sudo bash &&
        helm completion bash | sudo tee /etc/bash_completion.d/helm
        helm version
    fi"

    # Install kustomize
    ##VERSION## https://github.com/kubernetes-sigs/kustomize/releases
    KUSTOMIZE_VERSION=v5.4.1
    URL=https://github.com/kubernetes-sigs/kustomize/releases/download/kustomize/${KUSTOMIZE_VERSION}/kustomize_${KUSTOMIZE_VERSION}_linux_${ARCH}.tar.gz
    pssh "
    if [ ! -x /usr/local/bin/kustomize ]; then
        curl -fsSL $URL |
        sudo tar -C /usr/local/bin -zx kustomize
        kustomize completion bash | sudo tee /etc/bash_completion.d/kustomize
        kustomize version
    fi"

    # Install ship
    # Note: 0.51.3 is the last version that doesn't display GIN-debug messages
    # (don't want to get folks confused by that!)
    # Only install ship on Intel platforms (no ARM 64 builds).
    [ "$ARCH" = "amd64" ] &&
    pssh "
    if [ ! -x /usr/local/bin/ship ]; then
        ##VERSION##
        curl -fsSL https://github.com/replicatedhq/ship/releases/download/v0.51.3/ship_0.51.3_linux_$ARCH.tar.gz |
            sudo tar -C /usr/local/bin -zx ship
    fi"

    # Install the AWS IAM authenticator
    pssh "
    if [ ! -x /usr/local/bin/aws-iam-authenticator ]; then
        ##VERSION##
        sudo curl -fsSLo /usr/local/bin/aws-iam-authenticator https://amazon-eks.s3-us-west-2.amazonaws.com/1.12.7/2019-03-27/bin/linux/$ARCH/aws-iam-authenticator
	      sudo chmod +x /usr/local/bin/aws-iam-authenticator
        aws-iam-authenticator version
    fi"

    # Install jless (jless.io)
    pssh "
    if [ ! -x /usr/local/bin/jless ]; then
        ##VERSION##
        sudo apt-get install -y libxcb-render0 libxcb-shape0 libxcb-xfixes0
        wget https://github.com/PaulJuliusMartinez/jless/releases/download/v0.9.0/jless-v0.9.0-x86_64-unknown-linux-gnu.zip
        unzip jless-v0.9.0-x86_64-unknown-linux-gnu
        sudo mv jless /usr/local/bin
    fi"

    # Install the krew package manager
    pssh "
    if [ ! -d /home/$USER_LOGIN/.krew ]; then
        cd /tmp &&
        KREW=krew-linux_$ARCH
        curl -fsSL https://github.com/kubernetes-sigs/krew/releases/latest/download/\$KREW.tar.gz |
        tar -zxf- &&
        sudo -u $USER_LOGIN -H ./\$KREW install krew &&
        echo export PATH=/home/$USER_LOGIN/.krew/bin:\\\$PATH | sudo -u $USER_LOGIN tee -a /home/$USER_LOGIN/.bashrc
    fi"

    # Install kubecolor
    KUBECOLOR_VERSION=0.4.0
    URL=https://github.com/kubecolor/kubecolor/releases/download/v${KUBECOLOR_VERSION}/kubecolor_${KUBECOLOR_VERSION}_linux_${ARCH}.tar.gz
    pssh "
    if [ ! -x /usr/local/bin/kubecolor ]; then
        ##VERSION##
        curl -fsSL $URL |
        sudo tar -C /usr/local/bin -zx kubecolor
    fi"

    # Install k9s
    pssh "
    if [ ! -x /usr/local/bin/k9s ]; then
        FILENAME=k9s_Linux_$ARCH.tar.gz &&
        curl -fsSL https://github.com/derailed/k9s/releases/latest/download/\$FILENAME |
        sudo tar -C /usr/local/bin -zx k9s
        k9s version
    fi"

    # Install popeye
    pssh "
    if [ ! -x /usr/local/bin/popeye ]; then
        FILENAME=popeye_Linux_$ARCH.tar.gz &&
        curl -fsSL https://github.com/derailed/popeye/releases/latest/download/\$FILENAME |
        sudo tar -C /usr/local/bin -zx popeye
        popeye version
    fi"

    # Install Tilt
    # Official instructions:
    # curl -fsSL https://raw.githubusercontent.com/tilt-dev/tilt/master/scripts/install.sh | bash
    # But the install script is not arch-aware (see https://github.com/tilt-dev/tilt/pull/5050).
    pssh "
    if [ ! -x /usr/local/bin/tilt ]; then
        TILT_VERSION=0.33.13
        FILENAME=tilt.\$TILT_VERSION.linux.$TILT_ARCH.tar.gz
        curl -fsSL https://github.com/tilt-dev/tilt/releases/download/v\$TILT_VERSION/\$FILENAME |
        sudo tar -C /usr/local/bin -zx tilt
        tilt completion bash | sudo tee /etc/bash_completion.d/tilt
        tilt version
    fi"

    # Install Skaffold
    pssh "
    if [ ! -x /usr/local/bin/skaffold ]; then
        curl -fsSLo skaffold https://storage.googleapis.com/skaffold/releases/latest/skaffold-linux-$ARCH &&
        sudo install skaffold /usr/local/bin/
        skaffold completion bash | sudo tee /etc/bash_completion.d/skaffold
        skaffold version
    fi"

    # Install Kompose
    pssh "
    if [ ! -x /usr/local/bin/kompose ]; then
        curl -fsSLo kompose https://github.com/kubernetes/kompose/releases/latest/download/kompose-linux-$ARCH &&
        sudo install kompose /usr/local/bin
        kompose completion bash | sudo tee /etc/bash_completion.d/kompose
        kompose version
    fi"

    # Install KinD
    pssh "
    if [ ! -x /usr/local/bin/kind ]; then
        curl -fsSLo kind https://github.com/kubernetes-sigs/kind/releases/latest/download/kind-linux-$ARCH &&
        sudo install kind /usr/local/bin
        kind completion bash | sudo tee /etc/bash_completion.d/kind
        kind version
    fi"

    # Install YTT
    pssh "
    if [ ! -x /usr/local/bin/ytt ]; then
        curl -fsSLo ytt https://github.com/vmware-tanzu/carvel-ytt/releases/latest/download/ytt-linux-$ARCH &&
        sudo install ytt /usr/local/bin
        ytt completion bash | sudo tee /etc/bash_completion.d/ytt
        ytt version
    fi"

    ##VERSION## https://github.com/bitnami-labs/sealed-secrets/releases
    KUBESEAL_VERSION=0.26.2
    URL=https://github.com/bitnami-labs/sealed-secrets/releases/download/v${KUBESEAL_VERSION}/kubeseal-${KUBESEAL_VERSION}-linux-${ARCH}.tar.gz
    #case $ARCH in
    #amd64) FILENAME=kubeseal-linux-amd64;;
    #arm64) FILENAME=kubeseal-arm64;;
    #*)     FILENAME=nope;;
    #esac
    pssh "
    if [ ! -x /usr/local/bin/kubeseal ]; then
        curl -fsSL $URL |
        sudo tar -C /usr/local/bin -zx kubeseal
        kubeseal --version
    fi"

    ##VERSION## https://github.com/vmware-tanzu/velero/releases
    VELERO_VERSION=1.13.2
    pssh "
    if [ ! -x /usr/local/bin/velero ]; then
        curl -fsSL https://github.com/vmware-tanzu/velero/releases/download/v$VELERO_VERSION/velero-v$VELERO_VERSION-linux-$ARCH.tar.gz |
        sudo tar --strip-components=1 --wildcards -zx -C /usr/local/bin '*/velero'
        velero completion bash | sudo tee /etc/bash_completion.d/velero
        velero version --client-only
    fi"

    ##VERSION## https://github.com/doitintl/kube-no-trouble/releases
    KUBENT_VERSION=0.7.2
    pssh "
    if [ ! -x /usr/local/bin/kubent ]; then
        curl -fsSL https://github.com/doitintl/kube-no-trouble/releases/download/${KUBENT_VERSION}/kubent-${KUBENT_VERSION}-linux-$ARCH.tar.gz |
        sudo tar -zxvf- -C /usr/local/bin kubent
        kubent --version
    fi"

    # Ngrok. Note that unfortunately, this is the x86_64 binary.
    # We might have to rethink how to handle this for multi-arch environments.
    pssh "
    if [ ! -x /usr/local/bin/ngrok ]; then
        curl -fsSL https://bin.equinox.io/c/bNyj1mQVY4c/ngrok-v3-stable-linux-amd64.tgz |
        sudo tar -zxvf- -C /usr/local/bin ngrok
    fi"
}

_cmd kubereset "Wipe out Kubernetes configuration on all nodes"
_cmd_kubereset() {
    TAG=$1
    need_tag

    pssh "sudo kubeadm reset --force"
}

_cmd kubetest "Check that all nodes are reporting as Ready"
_cmd_kubetest() {
    TAG=$1
    need_tag

    # There are way too many backslashes in the command below.
    # Feel free to make that better ♥
    pssh "
    set -e
    if i_am_first_node; then
      which kubectl
      for NODE in \$(grep [0-9]\$ /etc/hosts | grep -v ^127 | awk {print\ \\\$2}); do
        echo \$NODE ; kubectl get nodes | grep -w \$NODE | grep -w Ready
      done
    fi"
    echo kube_ok > tags/$TAG/status
}

_cmd ips "Show the IP addresses for a given tag"
_cmd_ips() {
    TAG=$1
    need_tag $TAG

    while true; do
        for I in $(seq $CLUSTERSIZE); do
            read ip || return 0
            printf "%s\t" "$ip"
        done
        printf "\n"
    done < tags/$TAG/ips.txt
}

_cmd inventory "List all VMs on a given provider (or across all providers if no arg given)"
_cmd_inventory() {
    FIXME
}

_cmd logins "Show login information for a group of instances"
_cmd_logins() {
    TAG=$1
    need_tag $TAG

    cat tags/$TAG/logins.jsonl \
    | jq -r '"\(if .codeServerPort then "\(.codeServerPort)\t" else "" end )\(.password)\tssh -l \(.login)\(if .port then " -p \(.port)" else "" end)\t\(.ipaddrs)"'
}

_cmd maketag "Generate a quasi-unique tag for a group of instances"
_cmd_maketag() {
    if [ -z $USER ]; then
        export USER=anonymous
    fi
    MS=$(($(date +%N | tr -d 0)/1000000))
    date +%Y-%m-%d-%H-%M-$MS-$USER
}

_cmd netfix "Disable GRO and run a pinger job on the VMs"
_cmd_netfix () {
    TAG=$1
    need_tag

    pssh "
    sudo ethtool -K ens3 gro off
    sudo tee /root/pinger.service <<EOF
[Unit]
Description=pinger

[Install]
WantedBy=multi-user.target

[Service]
WorkingDirectory=/
ExecStart=/bin/ping -w60 1.1
User=nobody
Group=nogroup
Restart=always
EOF
    sudo systemctl enable /root/pinger.service
    sudo systemctl start pinger"
}

_cmd ping "Ping VMs in a given tag, to check that they have network access"
_cmd_ping() {
    TAG=$1
    need_tag

    fping < tags/$TAG/ips.txt
}

_cmd stage2 "Finalize the setup of managed Kubernetes clusters"
_cmd_stage2() {
    TAG=$1
    need_tag

    cd tags/$TAG/stage2
    terraform init -upgrade
    terraform apply -auto-approve
    terraform output -raw logins_jsonl > ../logins.jsonl
    terraform output -raw ips_txt > ../ips.txt
    echo "stage2_ok" > status
}

_cmd standardize "Deal with non-standard Ubuntu cloud images"
_cmd_standardize() {
    TAG=$1
    need_tag

    # Try to log in as root.
    # If successful, make sure than we have:
    # - sudo
    # - ubuntu user
    # Note that on Scaleway, the keys of the root account get copied
    # a little bit later after boot; so the first time we run "standardize"
    # we might end up copying an incomplete authorized_keys file.
    # That's why we copy it inconditionally here, rather than checking
    # for existence and skipping if it already exists.
    pssh -l root -t 5 true 2>&1 >/dev/null && {
        pssh -l root "
        grep DEBIAN_FRONTEND /etc/environment || echo DEBIAN_FRONTEND=noninteractive >> /etc/environment
        #grep cloud-init /etc/sudoers && rm /etc/sudoers
        apt-get update && apt-get install sudo -y
        getent passwd ubuntu || {
            useradd ubuntu -m -s /bin/bash
            echo 'ubuntu ALL=(ALL:ALL) NOPASSWD:ALL' > /etc/sudoers.d/ubuntu
        }
        install --owner=ubuntu --mode=700 --directory /home/ubuntu/.ssh
        install --owner=ubuntu --mode=600 /root/.ssh/authorized_keys --target-directory /home/ubuntu/.ssh
        "
    }

    # Now make sure that we have an ubuntu user
    pssh true

    # Disable unattended upgrades so that they don't mess up with the subsequent steps
    pssh sudo rm -f /etc/apt/apt.conf.d/50unattended-upgrades

    # Some cloud providers think that it's smart to disable password authentication.
    # We need to re-neable it, though.
    # Digital Ocecan
    pssh "
    if [ -f /etc/ssh/sshd_config.d/50-cloud-init.conf ]; then
        sudo rm /etc/ssh/sshd_config.d/50-cloud-init.conf
        sudo systemctl restart ssh.service
    fi"
    # AWS
    pssh "if [ -f /etc/ssh/sshd_config.d/60-cloudimg-settings.conf ]; then
        sudo rm /etc/ssh/sshd_config.d/60-cloudimg-settings.conf
        sudo systemctl restart ssh.service
    fi"

    # Special case for oracle since their iptables blocks everything but SSH
    pssh "
    if [ -f /etc/iptables/rules.v4 ]; then
        sudo sed -i 's/-A INPUT -j REJECT --reject-with icmp-host-prohibited//' /etc/iptables/rules.v4
        sudo netfilter-persistent flush
        sudo netfilter-persistent start
    fi"

    # oracle-cloud-agent upgrades pacakges in the background.
    # This breaks our deployment scripts, because when we invoke apt-get, it complains
    # that the lock already exists (symptom: random "Exited with error code 100").
    # Workaround: if we detect oracle-cloud-agent, remove it.
    # But this agent seems to also take care of installing/upgrading
    # the unified-monitoring-agent package, so when we stop the snap,
    # it can leave dpkg in a broken state. We "fix" it with the 2nd command.
    pssh "
    if [ -d /snap/oracle-cloud-agent ]; then
        sudo snap remove oracle-cloud-agent
        sudo dpkg --remove --force-remove-reinstreq unified-monitoring-agent
    fi"
}

_cmd tailhist "Install history viewer on port 1088"
_cmd_tailhist () {
    TAG=$1
    need_tag

    ARCH=${ARCHITECTURE-amd64}
    [ "$ARCH" = "aarch64" ] && ARCH=arm64

    # We use "wget -c" here in case the download was aborted
    # halfway through and we're actually trying to download it again.
    pssh "
    set -e
    sudo apt-get install unzip -y
    wget -c https://github.com/joewalnes/websocketd/releases/download/v0.3.0/websocketd-0.3.0-linux_$ARCH.zip
    unzip -o websocketd-0.3.0-linux_$ARCH.zip websocketd
    sudo mv websocketd /usr/local/bin/websocketd
    sudo mkdir -p /opt/tailhist
    sudo tee /opt/tailhist.service <<EOF
[Unit]
Description=tailhist

[Install]
WantedBy=multi-user.target

[Service]
WorkingDirectory=/opt/tailhist
ExecStart=/usr/local/bin/websocketd --port=1088 --staticdir=. sh -c \"tail -n +1 -f /home/$USER_LOGIN/.history || echo 'Could not read history file. Perhaps you need to \\\"chmod +r .history\\\"?'\"
User=nobody
Group=nogroup
Restart=always
EOF
    sudo systemctl enable /opt/tailhist.service --now
    "

    pssh -I sudo tee /opt/tailhist/index.html <lib/tailhist.html
}

_cmd terraform "Apply Terraform configuration to provision resources."
_cmd_terraform() {
    TAG=$1
    need_tag
    echo terraforming > tags/$TAG/status
    (
        cd tags/$TAG
        terraform apply -auto-approve
        # The Terraform provider for Proxmox has a bug; sometimes it fails
        # to obtain VM address from the QEMU agent. In that case, we put
        # ERROR in the ips.txt file (instead of the VM IP address). Detect
        # that so that we run Terraform again (this typically solves the issue).
        if grep -q ERROR ips.txt; then
          die "Couldn't obtain IP address of some machines. Try to re-run terraform."
        fi
    )
    echo terraformed > tags/$TAG/status

}

_cmd tools "Install a bunch of useful tools (editors, git, jq...)"
_cmd_tools() {
    TAG=$1
    need_tag

    pssh "
    set -e
    sudo apt-get -q update
    sudo apt-get -qy install apache2-utils argon2 emacs-nox git httping htop jid joe jq mosh tree unzip
    # This is for VMs with broken PRNG (symptom: running docker-compose randomly hangs)
    sudo apt-get -qy install haveged
    "
}

_cmd pssh "Run an arbitrary command on all nodes"
_cmd_pssh() {
    TAG=$1
    need_tag
    shift

    pssh "$@"
}

_cmd pull_images "Pre-pull a bunch of Docker images"
_cmd_pull_images() {
    TAG=$1
    need_tag
    pull_tag
}

_cmd remap_nodeports "Remap NodePort range to 10000-10999"
_cmd_remap_nodeports() {
    TAG=$1
    need_tag

    FIND_LINE="    - --service-cluster-ip-range=10.96.0.0\/12"
    ADD_LINE="    - --service-node-port-range=10000-10999"
    MANIFEST_FILE=/etc/kubernetes/manifests/kube-apiserver.yaml
    pssh "
    if i_am_first_node && ! grep -q '$ADD_LINE' $MANIFEST_FILE; then
        sudo sed -i 's/\($FIND_LINE\)\$/\1\n$ADD_LINE/' $MANIFEST_FILE
    fi"

    info "If you have manifests hard-coding nodePort values,"
    info "you might want to patch them with a command like:"
    info "

if i_am_first_node; then
    kubectl -n kube-system patch svc prometheus-server \\
        -p 'spec: { ports: [ {port: 80, nodePort: 10101} ]}'
fi

    "
}

_cmd ssh "Open an SSH session to the first node of a tag"
_cmd_ssh() {
    TAG=$1
    need_tag
    IP=$(head -1 tags/$TAG/ips.txt)
    info "Logging into $IP (default password: $USER_PASSWORD)"
    ssh $SSHOPTS $USER_LOGIN@$IP

}

_cmd tags "List groups of VMs known locally"
_cmd_tags() {
    (
        cd tags
        echo "[#] [Status] [Tag] [Mode] [Provider]"
        for tag in *; do
            if [ -f $tag/logins.jsonl ]; then
                count="$(wc -l < $tag/logins.jsonl)"
            else
                count="?"
            fi
            if [ -f $tag/status ]; then
                status="$(cat $tag/status)"
            else
                status="?"
            fi
            if [ -f $tag/mode ]; then
                mode="$(cat $tag/mode)"
            else
                mode="?"
            fi
            if [ -f $tag/provider ]; then
                provider="$(cat $tag/provider)"
            else
                provider="?"
            fi
            echo "$count $status $tag $mode $provider"
        done
    ) | column -t
}

_cmd test "Run tests (pre-flight checks) on a group of VMs"
_cmd_test() {
    TAG=$1
    need_tag
    test_tag
}

_cmd tmux "Log into the first node and start a tmux server"
_cmd_tmux() {
    TAG=$1
    need_tag
    IP=$(head -1 tags/$TAG/ips.txt)
    info "Opening ssh+tmux with $IP"
    rm -f /tmp/tmux-$UID/default
    ssh $SSHOPTS -t -L /tmp/tmux-$UID/default:/tmp/tmux-1001/default docker@$IP tmux new-session -As 0
}

_cmd helmprom "Install Prometheus with Helm"
_cmd_helmprom() {
    TAG=$1
    need_tag
    pssh "
    if i_am_first_node; then
        sudo -u $USER_LOGIN -H helm upgrade --install prometheus prometheus \
            --repo https://prometheus-community.github.io/helm-charts/ \
            --namespace prometheus --create-namespace \
            --set server.service.type=NodePort \
            --set server.service.nodePort=30090 \
            --set server.persistentVolume.enabled=false \
            --set alertmanager.enabled=false
    fi"
}

_cmd passwords "Set individual passwords for each cluster"
_cmd_passwords() {
    TAG=$1
    need_tag
    PASSWORDS_FILE="tags/$TAG/passwords"
    if ! [ -f "$PASSWORDS_FILE" ]; then
        error "File $PASSWORDS_FILE not found. Please create it first."
        error "It should contain one password per line."
        error "It should have as many lines as there are clusters."
        die "Aborting."
    fi
    N_CLUSTERS=$($0 ips "$TAG" | wc -l)
    N_PASSWORDS=$(wc -l < "$PASSWORDS_FILE")
    if [ "$N_CLUSTERS" != "$N_PASSWORDS" ]; then
        die "Found $N_CLUSTERS clusters and $N_PASSWORDS passwords. Aborting."
    fi
    $0 ips "$TAG" | paste "$PASSWORDS_FILE" - | while read password nodes; do
        info "Setting password for $nodes..."
        for node in $nodes; do
            echo $USER_LOGIN $password | ssh $SSHOPTS -i tags/$TAG/id_rsa ubuntu@$node '
                read login password
                echo $login:$password | sudo chpasswd
                hashedpassword=$(echo -n $password | argon2 saltysalt$RANDOM -e)
                sudo -u $login mkdir -p /home/$login/.config/code-server
                echo "hashed-password: \"$hashedpassword\"" | sudo -u $login tee /home/$login/.config/code-server/config.yaml >/dev/null
                '
        done
    done
    info "Done."
}

_cmd wait "Wait until VMs are ready (reachable, cloud init is done, ubuntu user is up)"
_cmd_wait() {
    TAG=$1
    need_tag

    # Wait until all hosts are reachable.
    info "Trying to reach $TAG instances..."
    while >/dev/stderr echo -n "."; do
        pssh -t 5 true 2>&1 >/dev/null && {
            SSH_USER=ubuntu
            break
        }
        pssh -l root -t 5 true 2>&1 >/dev/null && {
            SSH_USER=root
            break
        }
        sleep 2
    done
    >/dev/stderr echo ""

    # If this VM image is using cloud-init,
    # wait for cloud-init to be done
    info "Waiting for cloud-init to be done on $TAG instances..."
    pssh -l $SSH_USER "
    if [ -d /var/lib/cloud ]; then
        cloud-init status --wait
        case $? in
        0) exit 0;; # all is good
        2) exit 0;; # recoverable error (happens with proxmox deprecated cloud-init payloads)
        *) exit 1;; # all other problems
        esac
    fi"
}

# Sometimes, weave fails to come up on some nodes.
# Symptom: the pods on a node are unreachable (they don't even ping).
# Remedy: wipe out Weave state and delete weave pod on that node.
# Specifically, identify the weave pod that is defective, then:
# kubectl -n kube-system exec weave-net-XXXXX -c weave rm /weavedb/weave-netdata.db
# kubectl -n kube-system delete pod weave-net-XXXXX
_cmd weavetest "Check that weave seems properly setup"
_cmd_weavetest() {
    TAG=$1
    need_tag
    pssh "
    kubectl -n kube-system get pods -o name | grep weave | cut -d/ -f2 |
    xargs -I POD kubectl -n kube-system exec POD -c weave -- \
    sh -c \"./weave --local status | grep Connections | grep -q ' 1 failed' || ! echo POD \""
}

_cmd webssh "Install a WEB SSH server on the machines (port 1080)"
_cmd_webssh() {
    TAG=$1
    need_tag
    pssh "
    sudo apt-get update &&
    sudo apt-get install python3-tornado python3-paramiko -y"
    pssh "
    cd /opt
    [ -d webssh ] || sudo git clone https://github.com/jpetazzo/webssh"
    pssh "
    for KEYFILE in /etc/ssh/*.pub; do
      read a b c < \$KEYFILE; echo localhost \$a \$b
    done | sudo tee /opt/webssh/known_hosts"
    pssh "cat >webssh.service <<EOF
[Unit]
Description=webssh

[Install]
WantedBy=multi-user.target

[Service]
WorkingDirectory=/opt/webssh
ExecStart=/usr/bin/env python3 run.py --fbidhttp=false --port=1080 --policy=reject
User=nobody
Group=nogroup
Restart=always
EOF"
    pssh "
    sudo systemctl enable \$PWD/webssh.service &&
    sudo systemctl start webssh.service"
}

_cmd www "Run a web server to access card HTML and PDF"
_cmd_www() {
    cd www
    IPADDR=$(curl -fsSL canihazip.com/s || echo localhost)
    info "The following files are available:"
    for F in *; do
        echo "http://$IPADDR:8000/$F"
    done
    info "Press Ctrl-C to stop server."
    python3 -m http.server
}

pull_tag() {
    # Pre-pull a bunch of images
    pssh --timeout 900 'for I in \
        debian:latest \
        ubuntu:latest \
        fedora:latest \
        centos:latest \
        elasticsearch:2 \
        postgres \
        redis \
        alpine \
        registry \
        nicolaka/netshoot \
        jpetazzo/trainingwheels \
        golang \
        training/namer \
        dockercoins/hasher \
        dockercoins/rng \
        dockercoins/webui \
        dockercoins/worker \
        logstash \
        prom/node-exporter \
        google/cadvisor \
        dockersamples/visualizer \
        nathanleclaire/redisonrails; do
        sudo docker pull $I
    done'

    info "Finished pulling images for $TAG."
}

test_tag() {
    ips_file=tags/$TAG/ips.txt
    info "Picking a random IP address in $ips_file to run tests."
    ip=$(shuf -n1 $ips_file)
    test_vm $ip
    info "Tests complete."
}

test_vm() {
    ip=$1
    info "Testing instance with IP address $ip."
    user=ubuntu
    errors=""

    for cmd in "hostname" \
        "whoami" \
        "hostname -i" \
        "ls -l /usr/local/bin/i_am_first_node" \
        "grep . /etc/name_of_first_node /etc/ipv4_of_first_node" \
        "cat /etc/hosts" \
        "hostnamectl status" \
        "docker version | grep Version -B1" \
        "docker-compose version" \
        "docker-machine version" \
        "docker images" \
        "docker ps" \
        "curl --silent localhost:55555" \
        "sudo ls -la /mnt/ | grep docker" \
        "env" \
        "ls -la /home/docker/.ssh"; do
        sep "$cmd"
        echo "$cmd" \
            | ssh -A $SSHOPTS $user@$ip sudo -u docker -i \
            || {
                status=$?
                error "$cmd exit status: $status"
                errors="[$status] $cmd\n$errors"
            }
    done
    sep
    if [ -n "$errors" ]; then
        error "The following commands had non-zero exit codes:"
        printf "$errors"
    fi
    info "Test VM was $ip."
}

make_key_name() {
    SHORT_FINGERPRINT=$(ssh-add -l | grep RSA | head -n1 | cut -d " " -f 2 | tr -d : | cut -c 1-8)
    echo "${SHORT_FINGERPRINT}-${USER}"
}