garm/internal/templates/userdata/github_linux_userdata.tmpl

#!/bin/bash

set -e
set -o pipefail
set -x

{{- if .EnableBootDebug }}
set -x
{{- end }}

# Edit the templates directly in the browser. Intellisense is available for golang templates
# {{ .DownloadURL }}

CALLBACK_URL="{{ .CallbackURL }}"
METADATA_URL="{{ .MetadataURL }}"
BEARER_TOKEN="{{ .CallbackToken }}"

# touch /tmp/hold
# while true;do
#    [ -e /tmp/hold ] && sleep 2 || break
# done

RUN_HOME="/home/{{ .RunnerUsername }}/actions-runner"

if [ -z "$METADATA_URL" ];then
	echo "no token is available and METADATA_URL is not set"
	exit 1
fi

function call() {
	PAYLOAD="$1"
	[[ $CALLBACK_URL =~ ^(.*)/status(/)?$ ]] || CALLBACK_URL="${CALLBACK_URL}/status"
	curl --retry 5 --retry-delay 5 --retry-connrefused --fail -s -X POST -d "${PAYLOAD}" -H 'Accept: application/json' -H "Authorization: Bearer ${BEARER_TOKEN}" "${CALLBACK_URL}" || echo "failed to call home: exit code ($?)"
}

function systemInfo() {
	if [ -f "/etc/os-release" ];then
		. /etc/os-release
	fi
	OS_NAME=${NAME:-""}
	OS_VERSION=${VERSION_ID:-""}
	AGENT_ID=${1:-null}
	# strip status from the callback url
	[[ $CALLBACK_URL =~ ^(.*)/status(/)?$ ]] && CALLBACK_URL="${BASH_REMATCH[1]}" || true
	SYSINFO_URL="${CALLBACK_URL}/system-info/"
	PAYLOAD="{\"os_name\": \"$OS_NAME\", \"os_version\": \"$OS_VERSION\", \"agent_id\": $AGENT_ID}"
	curl --retry 5 --retry-delay 5 --retry-connrefused --fail -s -X POST -d "${PAYLOAD}" -H 'Accept: application/json' -H "Authorization: Bearer ${BEARER_TOKEN}" "${SYSINFO_URL}" || true
}

function sendStatus() {
	MSG="$1"
	call "{\"status\": \"installing\", \"message\": \"$MSG\"}"
}

function success() {
	MSG="$1"
	ID=${2:-null}
	call "{\"status\": \"idle\", \"message\": \"$MSG\", \"agent_id\": $ID}"
}

function fail() {
	MSG="$1"
	call "{\"status\": \"failed\", \"message\": \"$MSG\"}"
	exit 1
}

INSTANCE_METADATA=$(curl \
    --retry 5 \
    --retry-delay 5 \
    --retry-connrefused \
    --fail -s \
    -H 'Accept: application/json' \
    -H "Authorization: Bearer ${BEARER_TOKEN}" \
    "$METADATA_URL/runner-metadata" 2>&1) || fail "failed to get instance metadata: $INSTANCE_METADATA"

AGENT_MODE=$(echo "$INSTANCE_METADATA" | jq -r '.agent_mode // empty')
if [ "$AGENT_MODE" == "true" ]; then
    sendStatus "Agent mode is enabled; setting up agent"
    DOWNLOAD_URL=$(echo "$INSTANCE_METADATA" | jq -r '.agent_tools.download_url // empty')
    if [ -z "$DOWNLOAD_URL" ]; then
        fail "agent mode is enabled, but no agent tools found in metadata"
    fi
    AGENT_URL=$(echo "$INSTANCE_METADATA" | jq -r '.metadata_access.agent_url // empty')
    if [ -z "$AGENT_URL" ]; then
        fail "agent mode is enabled, but agent_url was not found in metadata"
    fi
    AGENT_TOKEN=$(echo "$INSTANCE_METADATA" | jq -r '.agent_token // empty')
    if [ -z "$AGENT_TOKEN" ]; then
        fail "agent mode is enabled, but agent_token was not found in metadata"
    fi
    AGENT_SHELL=$(echo "$INSTANCE_METADATA" | jq -r '.agent_shell_enabled // false')
    sendStatus "Downloading agent from $DOWNLOAD_URL"
    sudo curl --retry 5 \
        --retry-delay 5 \
        --retry-connrefused \
        --fail -L \
        -H "Authorization: Bearer ${BEARER_TOKEN}" \
        -o /usr/local/bin/garm-agent "$DOWNLOAD_URL" || fail "failed to download garm-agent"
    sudo chmod +x /usr/local/bin/garm-agent || fail "failed to make garm-agent executable"
    sudo mkdir -p /var/log/garm-agent || fail "failed to create /var/log/garm-agent"
    sudo chown {{ .RunnerUsername }}:{{ .RunnerUsername }} /var/log/garm-agent || fail "failed to chown /var/log/garm-agent"
    sudo mkdir -p /etc/garm-agent || fail "failed to create /etc/garm"
    sudo chown {{ .RunnerUsername }}:{{ .RunnerUsername }} /etc/garm-agent || fail "failed to change owner on /etc/garm-agent"

    sendStatus "Creating config and systemd unit"
    cat > /etc/garm-agent/garm-agent.toml << EOF
server_url = "$AGENT_URL"
log_file = "/var/log/garm-agent/garm-agent.log"
work_dir = "$RUN_HOME"
enable_shell = $AGENT_SHELL
token = "$AGENT_TOKEN"
runner_cmdline = ["/bin/bash", "-C", "/home/runner/actions-runner/run.sh"]
state_db_path = "/etc/garm-agent/agent-state.db"
EOF

    cat > /tmp/garm-agent.service << EOF
[Unit]
Description=GARM agent
After=multi-user.target

[Service]
Type=simple
ExecStart=/usr/local/bin/garm-agent daemon --config /etc/garm-agent/garm-agent.toml
ExecReload=/bin/kill -HUP $MAINPID
Restart=always
RestartSec=5s
User={{ .RunnerUsername }}
Environment=TERM=xterm-256color
Environment=LANG=en_US.UTF-8

[Install]
WantedBy=multi-user.target
EOF

    sudo mv /tmp/garm-agent.service /etc/systemd/system/garm-agent.service || fail "failed to create /etc/systemd/system/garm-agent.service"
    sudo chown root:root /etc/systemd/system/garm-agent.service || fail "failed to change owner on /etc/systemd/system/garm-agent.service"
    sendStatus "Reloading systemd unit"
    sudo systemctl daemon-reload || fail "failed to reload systemd"
fi

function downloadAndExtractRunner() {
	sendStatus "downloading tools from {{ .DownloadURL }}"
	if [ ! -z "{{ .TempDownloadToken }}" ]; then
	TEMP_TOKEN="Authorization: Bearer {{ .TempDownloadToken }}"
	fi
	curl --retry 5 --retry-delay 5 --retry-connrefused --fail -L -H "${TEMP_TOKEN}" -o "/home/{{ .RunnerUsername }}/{{ .FileName }}" "{{ .DownloadURL }}" || fail "failed to download tools"
	mkdir -p "$RUN_HOME" || fail "failed to create actions-runner folder"
	sendStatus "extracting runner"
	tar xf "/home/{{ .RunnerUsername }}/{{ .FileName }}" -C "$RUN_HOME"/ || fail "failed to extract runner"
	chown {{ .RunnerUsername }}:{{ .RunnerGroup }} -R "$RUN_HOME"/ || fail "failed to change owner"
}

if [ ! -d "$RUN_HOME" ];then
	downloadAndExtractRunner
	sendStatus "installing dependencies"
	cd "$RUN_HOME"
	attempt=1
	while true; do
		sudo ./bin/installdependencies.sh && break
		if [ $attempt -gt 5 ];then
			fail "failed to install dependencies after $attempt attempts"
		fi
		sendStatus "failed to install dependencies (attempt $attempt): (retrying in 15 seconds)"
		attempt=$((attempt+1))
		sleep 15
	done
else
	sendStatus "using cached runner found in $RUN_HOME"
	cd "$RUN_HOME"
fi


sendStatus "configuring runner"
{{- if .UseJITConfig }}
function getRunnerFile() {
	curl --retry 5 --retry-delay 5 \
		--retry-connrefused --fail -s \
		-X GET -H 'Accept: application/json' \
		-H "Authorization: Bearer ${BEARER_TOKEN}" \
		"${METADATA_URL}/$1" -o "$2"
}

sendStatus "downloading JIT credentials"
getRunnerFile "credentials/runner" ""$RUN_HOME"/.runner" || fail "failed to get runner file"
getRunnerFile "credentials/credentials" ""$RUN_HOME"/.credentials" || fail "failed to get credentials file"
getRunnerFile "credentials/credentials_rsaparams" ""$RUN_HOME"/.credentials_rsaparams" || fail "failed to get credentials_rsaparams file"
getRunnerFile "system/service-name" ""$RUN_HOME"/.service" || fail "failed to get service name file"
sed -i 's/$/\.service/' "$RUN_HOME"/.service

SVC_NAME=$(cat "$RUN_HOME"/.service)

sendStatus "generating systemd unit file"
getRunnerFile "systemd/unit-file?runAsUser={{ .RunnerUsername }}" "$SVC_NAME" || fail "failed to get service file"
sudo mv $SVC_NAME /etc/systemd/system/ || fail "failed to move service file"
sudo chown root:root /etc/systemd/system/$SVC_NAME || fail "failed to change owner"
if [ -e "/sys/fs/selinux" ];then
	sudo chcon -h system_u:object_r:systemd_unit_file_t:s0 /etc/systemd/system/$SVC_NAME || fail "failed to change selinux context"
fi

sendStatus "enabling runner service"
cp "$RUN_HOME"/bin/runsvc.sh "$RUN_HOME"/ || fail "failed to copy runsvc.sh"
sudo systemctl daemon-reload || fail "failed to reload systemd"
sudo systemctl enable $SVC_NAME
{{- else}}

GITHUB_TOKEN=$(curl --retry 5 --retry-delay 5 --retry-connrefused --fail -s -X GET -H 'Accept: application/json' -H "Authorization: Bearer ${BEARER_TOKEN}" "${METADATA_URL}/runner-registration-token/")


set +e
attempt=1
while true; do
	ERROUT=$(mktemp)
	{{- if .GitHubRunnerGroup }}
	./config.sh --unattended --url "{{ .RepoURL }}" --token "$GITHUB_TOKEN" --runnergroup {{.GitHubRunnerGroup}} --name "{{ .RunnerName }}" --labels "{{ .RunnerLabels }}" --no-default-labels --ephemeral 2>$ERROUT
	{{- else}}
	./config.sh --unattended --url "{{ .RepoURL }}" --token "$GITHUB_TOKEN" --name "{{ .RunnerName }}" --labels "{{ .RunnerLabels }}" --no-default-labels --ephemeral 2>$ERROUT
	{{- end}}
	if [ $? -eq 0 ]; then
		rm $ERROUT || true
		sendStatus "runner successfully configured after $attempt attempt(s)"
		break
	fi
	LAST_ERR=$(cat $ERROUT)
	echo "$LAST_ERR"

	# if the runner is already configured, remove it and try again. In the past configuring a runner
	# managed to register it but timed out later, resulting in an error.
	./config.sh remove --token "$GITHUB_TOKEN" || true

	if [ $attempt -gt 5 ];then
		rm $ERROUT || true
		fail "failed to configure runner: $LAST_ERR"
	fi

	sendStatus "failed to configure runner (attempt $attempt): $LAST_ERR (retrying in 5 seconds)"
	attempt=$((attempt+1))
	rm $ERROUT || true
	sleep 5
done
set -e

sendStatus "installing runner service"
sudo ./svc.sh install {{ .RunnerUsername }} || fail "failed to install service"
{{- end}}

if [ -e "/sys/fs/selinux" ];then
	sudo chcon -R -h user_u:object_r:bin_t:s0 /home/runner/ || fail "failed to change selinux context"
fi

AGENT_ID=""
{{- if .UseJITConfig }}
if [ -f "$RUN_HOME/env.sh" ];then
	pushd $RUN_HOME
	source env.sh
	popd
fi
if [ "$AGENT_MODE" != "true" ]; then
    sudo systemctl start $SVC_NAME || fail "failed to start service"
fi
{{- else}}
if [ "$AGENT_MODE" != "true" ]; then
    sendStatus "starting service"
    sudo ./svc.sh start || fail "failed to start service"
fi

set +e
AGENT_ID=$(grep "agentId" "$RUN_HOME"/.runner |  tr -d -c 0-9)
if [ $? -ne 0 ];then
	fail "failed to get agent ID"
fi
set -e
{{- end}}
systemInfo $AGENT_ID
if [ "$AGENT_MODE" == "true" ]; then
  sendStatus "Starting garm-agent service"
  sudo systemctl enable --now garm-agent || sendStatus "failed to start garm-agent"
fi
success "runner successfully installed" $AGENT_ID