generate-backends.sh

#!/bin/bash

PATH=${PATH}:/usr/sbin:/usr/bin:/opt/aws/bin:/opt/aws/bin
HOME=/root # Fix bug/problem with cronie in Amazon Linux (HOME=/ there)
# BACKEND_LAYER="$(cat /etc/varnish/backend-layer)"

VARNISH_MAINFILE="/etc/varnish/autoscalinggroup.vcl"
VARNISH_BACKENDSFILE="/etc/varnish/backends.vcl"

BACKEND_LAYER="$(aws ec2 describe-tags  --filters "Name=resource-id,Values=$(curl -s http://169.254.169.254/latest/meta-data/instance-id)" "Name=key,Values=backend-layer" | jq -r ".Tags[0].Value")"
# BACKEND_LAYER="ASG-Apache"
if [ -z "$BACKEND_LAYER" ]
then
  logger --stderr --priority user.warning "Cannot retrieve backend layer name"
  exit 1
fi


BACKENDS_DEFS=""
BACKENDS_INIT="
sub backends_init {
	new vdir = directors.round_robin();
"

BACKEND_CONFIG="
    	.port = \"80\";
	.max_connections = 300; # That's it
	.probe = {
		#.url = \"/\"; # short easy way (GET /)
		# We prefer to only do a HEAD /
		.request =
			\"HEAD / HTTP/1.1\"
			\"Host: localhost\"
			\"Connection: close\";
		.interval = 5s; # check the health of each backend every 5 seconds
		.timeout = 1s; # timing out after 1 second.
		# If 3 out of the last 5 polls succeeded the backend is considered healthy, otherwise it will be marked as sick
		.window = 5;
		.threshold = 3;
		}
	.first_byte_timeout     = 300s;   # How long to wait before we receive a first byte from our backend?
	.connect_timeout        = 5s;     # How long to wait for a backend connection?
	.between_bytes_timeout  = 2s;     # How long to wait between bytes received from our backend?

"
INSTANCES=$(aws autoscaling describe-auto-scaling-groups --auto-scaling-group-names "${BACKEND_LAYER}" | jq -r ".AutoScalingGroups[].Instances[].InstanceId")
if [ -z "$INSTANCES" ]
then
  logger --stderr --priority user.warning "Error retrieving autoscaling group instances, or no instances in group"
  exit 1
fi

MD5SUM_STATEFILE="/var/lib/misc/generate-backends.md5sum"
# Check md5sum of this list vs the value stored in /var/lib/misc/generate-backends.md5sum
MD5SUM_NEW="$(echo $INSTANCES | sort |md5sum | cut -d " " -f 1)"
if [ -e $MD5SUM_STATEFILE ] ; then
	MD5SUM_OLD="$(< $MD5SUM_STATEFILE )"
	if [ "$MD5SUM_NEW" == "$MD5SUM_OLD" ] ; then
		# Nothing to do
		exit
	fi
fi

# We haven't exited, so continue to generate the file

for ID in $INSTANCES ; do
	IP="$(aws ec2 describe-instances --instance-ids $ID --query Reservations[].Instances[].PrivateIpAddress --output text )"
  if [ -z "$IP" ]
  then
    logger --stderr --priority user.warning "Failed to retrieve ip address for instance $ID"
    exit 1
  fi
	BACKEND_NAME="$(echo "${ID}_$IP" | tr -- -. __)"
	echo "$ID -> $IP ($BACKEND_NAME)"
	BACKENDS_DEFS+="backend $BACKEND_NAME {
	.host = \"${IP}\";
	$BACKEND_CONFIG
}
"
	BACKENDS_INIT+="
	vdir.add_backend($BACKEND_NAME);"
done
BACKENDS_INIT+="
}"

cat > /etc/varnish/backends.vcl <<EOF
$BACKENDS_DEFS
$BACKENDS_INIT
EOF


TIMESTAMP=$(date +%Y%m%d-%H%M%S)
# Load (& compile) new vcl
varnishadm vcl.load vcl-${TIMESTAMP} $VARNISH_MAINFILE
res=$?
if [ $res -ne 0 ]
then
  logger --stderr --priority user.error "Failed to compile new varnish file"
  exit 1
fi
# Switch active vcl to the new one
varnishadm vcl.use vcl-${TIMESTAMP}

# Update list of currently active instances
echo $MD5SUM_NEW >$MD5SUM_STATEFILE

# Scan and delete old vcls
for i in $(varnishadm vcl.list |egrep -v "^active" |awk '{print $3;}') ; do
	varnishadm vcl.discard "${i}"
done

LOG_MSG="$BACKEND_LAYER varnish vcl updated with $(echo "$INSTANCES" | wc --lines) backends ($(echo "$INSTANCES" |paste --serial --delimiter=, ))"

logger --stderr --priority user.notice "$LOG_MSG"